INSTRUCTOR'S MANUAL
for
Historical Linguistics: an Introduction
(Third Edition)
Lyle Campbell University of Hawai‘i Mānoa
© Massachusetts Institute of Technology 2013 All rights reserved. No part of this manual may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. Not for sale – This manual is available only to instructors who have adopted the accompanying textbook for course use. The manual is not available to students, or to those studying the book by themselves. Instructors may not sell or transfer the manual to others or make any of the solutions publicly accessible, on the Web or otherwise.
INSTRUCTOR'S MANUAL
Historical Linguistics: an Introduction
Table of Contents PREFACE Chapter 1 Introduction, Exercise and solutions 1
Chapter 2 Sound change, Exercise and solutions 10
Chapter 3 Borrowing, Exercise and solutions 26
Chapter 4 Analogical change, Exercise and solutions 55
Chapter 5 Comparative method, Exercise and solutions 58
Chapter 6 Language Classification, Exercise and solutions 90
Chapter 7 Models of Linguistic Change (No exercises)
Chapter 8 Internal reconstruction, Exercise and solutions 94
Chapter 9 Semantic and lexical change, Exercise and solutions 119
Chapter 10 Morphological Change (No exercises)
Chapter 11 Syntactic change, Exercise and solutions 123
Chapter 12 Language Contact (No exercises)
Chapter 13 Explanation (No exercises)
Chapter 14 Distant Genetic Relationship, Exercise and solutions 129
Chapter 15 Philology and writing, Exercise and solutions (No exercises)
Chapter 16 Linguistics prehistory, Exercise and solutions 141
Chapter 17 Quantitative Approaches to Historical Linguistics (No exercises)
PREFACE
The primary mission of this manual is to provide sample answers or possible solutions to the exercises in the Historical Linguistics: an Introduction, 3rd edition. For many of the exercises, of course, there is no definitive or unique single solution, and other answers than those given here may often be possible. The purpose here is to provide sample possible answers which indicate how the exercise might be approached and solutions sought.
For ease of access, the exercises with their questions are repeated here, followed by possible solutions.
The exercise instructions and examples are color-coded in BLACK. Solutions are in BLUE. Notes and incidental comments are in RED
EXERCISES FOR CHAPTER 1 INTRODUCTION
Exercise 1.1
This exercise is about attitudes towards language change.
1. Try to find letters to newspapers or columns in newspapers or magazines, or on blogs, which express which express opinions on the quality of English in use today and about changes that are taking place. What do you think they reveal about attitudes towards language change?
2. Ask your friends, family and associates what they think about language today; do they think it is changing, and if so, is it getting better or worse?
3. Find books or articles on ‘proper’ English (prescriptive grammar); do they reveal any attitude towards changes that are going on in today’s language?
4. Consider the many things that schoolteachers or school grammar books warn you against as being ‘wrong’ or ‘bad grammar’. Do any of these involve changes in the language?
5. Compare books on etiquette written recently with some written thirty years ago or more; find the sections which deal with appropriate ways of speaking and use of the language. What changes have taken place in the recommendations made then and now? Do these reveal anything about change in the language or in language use?
Anything the reader finds involving attitudes towards language change can contribute to answering the questions of this exercise. An internet search for “attitudes towards language change”, “language attitudes”, “proper language”, “improper language”, “correct language”, “language etiquette”, “correct grammar”, “bad grammar”, or “prescriptive grammar”, for example, should provide sources relevant to parts 1., 3., 4., and 5. of this exercise. For part 2., ask friends, family, and associates.
Exercise 1.2
Observe the language you hear around you, and think about any changes that are going on now or have taken place in your lifetime. For example, if you are old enough, you might observe that gay has changed its basic meaning: today it mostly means ‘homosexual’ although until recently it did not have this meaning, but rather meant only ‘happy, cheerful’. Slang changes at a rather fast rate; what observations might you make about recent slang versus earlier slang? Can you find examples of ongoing change in other areas of the language besides just vocabulary?
Any observations about what may appear to be language change taking place currently are appropriate as answers to this question. Examples of lexical change are the easiest to identify
1 for people not trained in linguistics; slang and obscenities typically offer examples. You might notice such things as the tendency to lose the subjunctive in English, as in I wish Gollum was smarter instead of I wish Gollum were smarter, or the use of is is where standard English would have only a single is, as in The problem is is that his feet are too big, rather than standard The problem is that his feet are too big – perhaps on analogy to sentences such as, What the problem is is that his feet are too big. Some people are able to notice differences in pronunciation of vowels in some regions that seem to reflect ongoing changes.
Exercise 1.3
Changes in spelling and occasional misspellings have been used to make inferences about changes in pronunciation. This can, of course, be misleading, since spelling conventions are sometimes used for other purposes than just to represent pronunciation. Try to find examples of recent differences in spelling or of misspellings and then try to imagine what they might mean, say, to future linguists looking back trying to determine what changed and when it changed. For example, you might compare the spelling lite with light, gonna with going to, wannabee with want to be, or alright and alot with all right and a lot respectively. In particular, variations in spellings can be very revealing; see if you can find examples which may suggest something about language change.
Any examples of spellings not generally considered standard or examples of frequent misspellings can be possible answers. An internet search of things such as “occasional spelling”, “bad spelling”, “spelling errors”, etc. will turn up numerous examples.
Exercise 1.4 Shakespeare
A number of examples from Shakespeare’s plays, written in the Early Modern English period, are presented here which illustrate differences from how the same thing would be said today. Think about each example and attempt to state what changes have taken place in the language that would account for the differences you see in the constructions mentioned in the headings, the negatives, auxiliary verbs and so on. For example, in the first one we see: Saw you the weird sisters? The modern English equivalent would be Did you see the weird sisters? Had the heading directed your attention to yes–no questions, you would attempt to state what change had taken place, from former saw you (with inversion from you saw) to the modern version which no longer involves inversion but requires a form of do (did you see) which was not utilized in Shakespeare’s version.
Treatment of negatives:
1. Saw you the weird sisters? . . . Came they not by you? (Macbeth IV, i) 2. I love thee not, therefore pursue me not (A Midsummer Night’s Dream II, 1, 188) 3. I know thee not, old man: fall to thy prayers (Henry V V, v) 4. Let not thy mother lose her prayers, Hamlet: I pray thee, stay with us; go not to Wittenberg (Hamlet I, ii)
2 5. But yet you draw not iron (AMidsummer Night’s Dream II, i, 196) 6. Tempt not too much the hatred of my spirit (A Midsummer Night’s Dream II, i, 211) 7. And I am sick when I look not on you (A Midsummer Night’s Dream II, i, 213) 8. I will not budge for no man’s pleasure (Romeo and Juliet III, i) 9. I cannot weep, nor answer have I none (Othello IV, ii) 10. I am not sorry neither (Othello V, ii).
Negatives: Auxiliary do is inserted with main verbs (not with forms of to be or modals); not (or its contraction, -n’t) is placed after an auxiliary, modal verb, or a form of the verb to be: 1. Came they not by you? > Didn’t they come by you? (Auxiliary do). 2. I love thee not, therefore pursue me not > I don’t [or do not] love you, therefore don’t [or do not] pursue me. 3. I know thee not ... > I don’t [or do not] know you ...
The former use of double negatives is no longer grammatical in Standard English: 8. I will not budge for no man’s pleasure > I will not budge for any man’s pleasure / I will budge for no man’s pleasure. 9. I cannot weep, nor answer have I none > I cannot weep, nor do I have any answer. 10. I am not sorry neither > I am not sorry either.
Treatment of auxiliary verbs:
1. Macduff is fled to England (Macbeth IV, i) = ‘has fled’ 2. The king himself is rode to view their battle (Henry V IV, iii) = ‘has ridden’ 3. Thou told’st me they were stolen into this wood (A Midsummer Night’s Dream II, i, 191) = ‘had stolen away/hidden’
Auxiliary verbs: Forms of to be as the auxiliary verb in constructions with main verbs of motion in the Present Perfect or Past Perfect have changed to corresponding forms to the auxiliary have: 1. Macduff is fled to England > Macduff has fled to England. 2. The king himself is rode to view their battle > The king himself has ridden to view their battle. 3. Thou told’st me they were stolen into this wood > You told me they had stolen [hidden] in the woods.
Treatment of comparatives and superlatives:
1. She comes more nearer earth than she was wont (Othello 5, 2) 2. This was the most unkindest cut of all (Julius Caesar 3, 2) 3. What worser place can I beg in your love (A Midsummer Night’s Dream II, i, 208)
Comparatives and superlatives: The “double” comparatives and superlatives permitted in Shakespeare’s time are no longer permitted, so there are no longer combinations of more ... -er, most ...-est, or worser:
3 1. She comes more nearer earth than she was wont > She comes nearer to earth than she was wont [in the habit of doing]. 2. This was the most unkindest cut of all > This was the unkindest cut of all. 3. What worser place can I beg in your love > What worse place can I beg in your love.
Differences in verb agreement inflections (endings on the verbs which agree with the subject):
1. The quality of mercy is not strain’d It droppeth as the gentle rain from heaven Upon the place beneath: it is twice blessed; It blesseth him that gives and him that takes (The Merchant of Venice IV, i) 2. The one I’ll slay, the other slayeth me (A Midsummer Night’s Dream II, i, 190) 3. O, it offends me to the soul to Hear a robostious periwig-pated fellow tear A passion to tatters (Hamlet III, i, 9–11) 4. And could of men distinguish, her election Hath seal’d thee for herself: for thou hast been As one, in suffering all, that suffers nothing (Hamlet III, i, 68–71)
Differences in verb agreement inflections: The -(e)th of 3rd Person agreement on verbs > -(e)s; and -(e)st of 2nd Person Singular was lost when you replaced thou as the 2nd Person Pronoun, singular and plual. 1. droppeth > drops; blesseth > blesses; 2. slayeth > slays; 4. hath > has; (thou) hast > (you) have.
Exercise 1.5 Chaucer
The following is a sample text of Middle English, from Chaucer c. 1380. It is presented three lines at a time: the first is from Chaucer’s text; the second is a word-by-word translation, with some of the relevant grammatical morphemes indicated; the third is a modern translation. Compare these lines and report the main changes you observe in morphology, syntax, semantics and lexical items. (Do not concern yourself with the changes in spelling or pronunciation.)
The Tale of Melibee, Geoffrey Chaucer (c. 1380)
Upon a day bifel that he for his desport is went into the feeldes hym to pleye. on one day befell that he for his pleasure is gone to the fields him to play. ‘One day it happened that for his pleasure he went to the fields to amuse himself.’ [NOTE: is went = Modern English ‘has gone’; with verbs of motion the auxiliary used was a form of the verb ‘to be’, where today it is with ‘to have’]
4
His wif and eek his doghter hath he laft inwith his hous, his wife and also his daughter has he left within his house, ‘His wife and his daughter also he left inside his house,’ [NOTE: wif = ‘wife, woman’] of which the dores wer-en faste y-shette. of which the doors were-Plural fast Past.Participle-shut ‘whose doors were shut fast.’
Thre of his old foos ha-n it espied, and sett-en laddres to the walles of his hous, three of his old foes have-Plural it spied, and set-Plural ladders to the walls of his house, ‘Three of his old enemies saw this, and set ladders to the walls of his house,’ and by wyndowes ben entred, and betten his wyf, and by windows had entered, and beaten his wife, ‘and entered by the windows, and beat his wife,’ [NOTE: ben entred = ‘have entered’, a verb of motion taking ‘to be’ as the auxiliary] and wounded his doghter with fyve mortal woundes in fyve sondry places – and wounded his daughter with five mortal wounds in five sundry places – ‘and wounded his daughter with five mortal wounds in five different places –’ this is to sey-n, in hir feet, in hir handes, in hir erys, in hir nose, and in hir mouth, – this is to say-Infinitive, in her feet, in her hands, in her ears, in her nose, and in her mouth, – ‘that is to say, in her feet, in her hands, in her ears, in her nose, and in her mouth –’ and left-en hir for deed, and went-en awey. and left-Plural her for dead, and went-Plural away. ‘and left her for dead, and went away.’ (Lass 1992: 25–6)
Morphological changes: hath > has Replacement of 3rd Person Present agreement marker -th by -s. wer-en > were ha-n > have sett-en > set left-en > left went-en > went -(e)n ‘Plural agreement suffix’ of verbs was lost. y-shette > shut y- ‘Past Participle prefix’ was lost. to sey-n > to say
5 -n ‘Infinitive suffix’ was lost. laft in hath he laft > left
Syntactic changes: Upon a day bifel that > Upon a day it befell that = One day it happened that There is not enough evidence in this set of data to see this change properly, but English had a rule that inverted subject and verb when something preceded them in the clause (as in more or less frozen collocations such as Along came Jones, Never had he witnessed such folly, etc.); in cases with auxiliaries or modals in a verb phrase, just the auxiliary or modal was inverted, not the main verb. Thus here, upon a day comes first, and because the clause that he for his desport is went into the feeldes hym to pleye is the subject of bifel, it underwent subject-verb inversion to give Upon a day bifel that he for his desport is went into the feeldes hym to pleye. is went > has/had gone ben entered > have/had entered Forms of to be as the auxiliary verb with main verbs of motion in Present Perfect or Past Perfect changed to corresponding forms to the auxiliary to have.
Lexical or Semantic changes: bifel [befell] > happened befall is archaic in Modern English, essentially replaced by happen (lexical change). desport > pleasure desport has been replaced in this context (lexical change). pleye ‘play’ in hym to pleye > to amuse himself (lexical change: play > amuse, semantic change: play in hym to pleye lost the sense of ‘amuse’, ‘to amuse himself’). eek > also (lexical change). sondry ‘sundry’ > various sundry is archaic in Modern English, mostly replaced by various, different (lexical change).
Exercise 1.6 Caxton
The text in this exercise is a sample of Early Modern English, from William Caxton, Eneydos (c. 1491). As in Exercise 1.5, three lines are presented: the first is from Caxton’s text; the second is a word-by-word translation, with some of the relevant grammatical morphemes indicated; the third is a more colloquial modern translation. Compare these lines and report the main changes you observe in morphology, syntax, semantics and lexical items. (Again, do not concern yourself with the changes in spelling or pronunciation beyond the most obvious ones.)
And that commyn englysshe that is spoken in one shyre varyeth from a nother. In so moche and that common English that is spoken in one shire varies from another. In so much ‘And the common English that is spoken in one county varies so much from [that spoken in] another. In so much’
6 that in my days happened that certayn marchauntes were in a ship in tamyse that in my days happened that certain merchants were in a ship in Thames ‘that in my time it happened that some merchants were in a ship on the Thames’ for to haue sayled ouer the see to zelande/ and for lacke of wynde thei taryed atte forlond; for to have sailed over the sea to Zeeland. And for lack of wind they tarried at.the coast; ‘to sail over the sea to Zeeland. And because there was no wind, they stayed at the coast’ [NOTE: Zeeland = a province in the Netherlands] and wente to land for to refreshe them And one of theym, named sheffelde a mercer and went to land for to refresh them. And one of them, named Sheffield, a mercer, ‘and they went on land to refresh themselves. And one of them, named Sheffield, a fabric- dealer,’ cam in to an hows and axed [aksed] for mete, and specyally he axyd after eggys. came into a house and asked for meat, and especially he asked after eggs. ‘came into a house and asked for food, and specifically he asked for “eggs”.’
And the goode wyf answerede. that she coude no frenshe. and the good woman answered that she could no French. ‘And the good woman answered that she knew no French.’
And the marchaunt was angry. for he also coude speke no frenshe. and the merchant was angry, for he also could speak no French, ‘And the merchant was angry, because he couldn’t speak any French either.’ [NOTE: coude = ‘was able to, knew (how to)’] but wolde haue hadde egges/ and she vnderstode hym not/ but would have had eggs; and she understood him not. ‘but he wanted to have eggs; and she did not understand him.’ [NOTE: wolde = ‘wanted’, the source of Modern English would]
And thenne at laste a nother sayd that he wolde haue eyren/ and then at last an other said that he would have eggs. ‘and then finally somebody else said that he wanted to have eggs.’ then the good wyf said that she understod him wel/ then the good woman said that she understood him well. ‘Then the good woman said that she understood him well.’ (Source of Caxton’s text: Fisher and Bornstein 1974: 186–7)
Morphological changes: varyeth > varies Replacement of 3rd Person Present agreement marker -th by -s.
7 NOTE: eyren (ey singular) was native English for eggs, but was replaced by Scandinavian egg(s), a borrowing. Eyren also has the -en ‘plural suffix’ which was later lost.
Syntactic changes: In so much that in my days happened that certayn marchauntes were .... > So much that in my day it happened that certain merchants were ... There is not enough evidence in these data to see the change properly, but English had a rule that inverted subject and verb when something preceded them in the clause (as in more or less frozen collocations such as Along came Jones, Never had he witnessed such folly, etc.), but in cases with auxiliaries or modals in the verb phrase just the auxiliary or modal was inverted, not the main verb. Thus here, with in my days at the beginning of the clause, that certayn marchauntes were ... happened ... underwent inversion to ... happened that certayn marchauntes were ... for to haue sayled > (in order) to have sailed [or to sail] for to refresche them > (in order) to refresh themselves The former for to infinitival complementizer was reduced to just to when the subject of the main verb and of the infinitive were the same (i.e. co-referential), for example, they wanted to refresh themselves, where the subject of wanted and of to refresh is they for both, but they wanted for John to refresh himself) when the subject of wanted (= they) is different from the subject of to refresh (= John). refreseh them > refresh themselves Addition of self in reflexives. could speke no frenshe > could not [couldn’t] speak any French vnderstode hym not > did not [didn’t] understand him Main verbs when negated came to take a form of do in do-support, and the negative came to be placed next to auxiliaries and modals, including the form of do inserted in do-support.
Lexical or semantic changes: shyre ‘county’ > county (lexical change). forlond ‘foreland’ > coastal promontory ‘Foreland’ now archaic (lexical change). taryed ‘tarried’ > stayed, delayed ‘Tarry’ now feels archaic (lexical change). mercer > textile dealer (lexical change). mete ‘food’ > meat Change in meaning from meat as ‘food’ generally to the narrower meaning of meat as ‘food from animal flesh’ (semantic change). wyf ‘woman’ > wife Change in meaning from wife as ‘woman’ generally to ‘married woman, female spouse’ (semantic change). coude in coude no frenshe > knew (no French) Change in meaning from can / could ‘know (how to)’ / ‘ knew (how to)’
8 Meaning narrowed as can / could was grammaticalized as a modal verb only. wolde ‘wanted’ > wanted Change in meaning from wolde (Past tense of will ‘want’). The meaning narrowed as will and wolde were grammaticalized as modals only.
9
EXERCISES FOR CHAPTER 2 SOUND CHANGE
Exercise 2.1 Sound change – Proto-Germanic to Old English
Compare the Proto-Germanic forms with their descendants in Old English and determine what sound changes involving vowels have taken place. Write out the sound change involved, and identify (by name) the kind of change found. Ignore changes involving second syllables. (NOTE: ī, ō, and ū, are long vowels.) Proto-Germanic Old English 1. *fimf fīf ‘five’ 2. *gans- gōs ‘goose’ 3. *grinst grīst ‘a grinding’ ‘grist’ 4. *hanh- hōh ‘heel, hock’ 5. *linθj(az)- līθe ‘mild, lithe’ 6. *munθ- mūθ ‘mouth’ 7. *tanθ- tōθ ‘tooth’ 8. *gang- gang ‘a going’ 9. *grind- grind ‘grind’ 10. *hlink- hlink ‘ridge, links’ 11. *hund- hund ‘dog, hound’ 12. *land- land ‘land’ 13. *sing- sing- ‘sing’ 14. *slink- slink- ‘slink’ 15. *sundan sund- ‘swimming, sea, sound’ 16. *swing- swing- ‘swing’ 17. *θingam θing- ‘assembly, (legal) case, thing’ 18. *wund- wund ‘a wound’
Write out the sound changes involved: (1) VN > V: /__ Fricative A vowel and following nasal become a lengthened version of the vowel when followed by a fricative (as in 1.-7.). The change does not take place when followed by any consonant that was not a fricative (as in 8.-18.). (2) ā > ō In 2., 4., and 7. original *an ends up as ō when followed by a fricative. We can assume that an > ā along with the other vowels in change (1) when before a fricative and then subsequently ā > ō. However, some might assume that the change was rather first an > ō before a fricative, and then change (1) applied to the remaining vowels. This is much less likely though not out of the realm of possibility.
Identify by name the kind of sound change: Compensatory lengthening.
10 Note, though relative chronology is not asked for, as stated here, Change 1 must precede Change 2, since it is Change 1 which produces the ā which is then subsequently raised and rounded in Change 2.
Exercise 2.2 Sound change – Sanskrit to Pali
Compare the Sanskrit forms with their descendants in later Pali; determine what sound changes have taken place. Write out the changes, and identify (by name) the kind of changes where possible. NOTE: Sanskrit s = [s], ś = [ʃ], ṣ = [ʂ]. Each set is in effect a separate sound change exercise, though some changes may be illustrated in the examples of more than one set.
Set I Sanskrit Pali 1. śaśa sasa ‘hare’ 2. kēśa kesa ‘hair’ 3. dēśa desa ‘country’ 4. dōṣa dosa ‘fault’ 5. dāṣa dasa ‘slave’ 6. śiṣya sissa ‘pupil’ 7. sasya sassa ‘grain’ ˙
Set II Sanskrit Pali 8. snāna sināna ‘bathing’ 9. sneha sineha ‘friendship’ 10. snihyati sinihyati ‘is fond of’ 11. snigdha siniddha ‘oily’
Set III Sanskrit Pali 12. āuṣadha ōsadha ‘herbs, medicine’ 13. kāuśika kōsika ‘owl’ 14. gaura gōra ‘pale’ 15. mauna mōna ‘silence’ 16. augha ōgha ‘flood’ 17. tāila tēla ‘oil’ 18. vāira vēra ‘enmity’ 19. śāila sēla ‘rocky’ 20. aikya ekka ‘oneness’
Set IV Sanskrit Pali 21. pariṣat parisā ‘assembly’ 22. matimant matimā ‘wise’
11 23. ārakāt ārakā ‘from afar’ 24. dharmāt dhammā ‘merit (ablative)’ 25. arthāt atthā ‘that is’ 26. bhagavant bhagavā ‘venerable’ 27. mitravant mittavā ‘having friends’ (Bhat 2001: 67, 68, 70, Masica 1991: 168)
Determine what sound changes have taken place. Write out the changes, and identify (by name) the kind of changes where possible.
(1) Merger: ś, ṣ > s (1.-7., 12., 18.-20.) (2) Loss of vowel length: V V̅ > V (2.-5.) [[NOTE: the macron (line above the vowel) goes over (above) the V ̅ ]] (3) r-assimilation: r > t /t __ (27.) (4) C-cluster assimilation: C1 > C2[unaspirated] /C2 __ (C1 not = h) (6.-7., 18., 25.) First, (3) changed tr > tt, eliminating the single case (in 27.) where in the consonant cluster the second consonant assimilates completely to that of the preceding consonant. Then, by (4) in all remaining consonant clusters, the second consonant assimilates completely to match the first consonant except that if the second consonant is a voiced aspirate, the first consonant assimilates in all other features, but does not become aspirated: ṣy, sy > ss (6.-7.); gdh > ddh (11.); ky > kk (20.); rm > mm (24.) (5) Epenthesis: Ø > i /#s __ Nasal (8.-11.) Epenthesis of i between s and a nasal. (6) Monophthongization: au, āu > ō, ai, āi > ē /__ CV (12.-19.) (7) Vowel shortening before a consonant cluster: ē > e /__ CC (20.) [[NOTE: there is only one example to illustrate this change, but let’s assume that with more data, we would encounter additonal examples.]] (8) Final consonant loss with compensatory lenghtening: Vt, Vnt > V̅ /__ # (21.-27.) [[NOTE: the macron (line above the vowel) goes over (above) the V ̅ ]] [Note that this change could be stated as involving only final alveolar consonants (instead of all final consonants) if one believes that additional data might reveal other, non-alveolar consonants in final position that do not get deleted leaving compensatory lengthening, since in these data t and nt are the only final consonants. Also, it is possible to think that the change took place in stages, first deleting final t (as in 26. bhagavant > bhagavan ‘venerable’) and then later when n showed up in final position because of this change, it too was deleted with compensatory lengthening ((26. bhagavant > bhagavā ‘venerable’)).
12 Exercise 2.3 Sound change – Sanskrit to Prakrit
Compare the Sanskrit forms with their descendants in later Prakrit; determine what sound changes have taken place. Write out the changes, and identify (by name) the kind of changes where possible. NOTE: consonants with subscript dots are retroflex; Sanskrit s = [s], ś = [ʃ], ṣ = [ʂ]. Each set is in effect a separate sound change exercise, though some changes may be illustrated in the examples of more than one set.
Set I Sanskrit Prakrit 1. sapta satta ‘seven’ 2. dugdha duddha ‘milk’ 3. udgāra uggāla ‘spit out’ 4. tikta titta ‘pungent’ 5. mudga mugga ‘mung bean’ 6. ardha addha ‘half’ 7. karpaṭa kappaḍa ‘rag, cloth’ 8. kurkura kukkura ‘dog’ 9. darpa dappa ‘arrogance’ 10. parṇa paṇṇa ‘leaf’ 11. karma kamma ‘work’
Set II Sanskrit Prakrit 12. saras sara ‘lake’ 13. śara sara ‘arrow’ 14. sapta satta ‘seven’ 15. śakta satta ‘able’ 16. sarva savva ‘all’ 17. śava savva ‘corpse’ 18. sīsa sīsa ‘lead’ 19. śīla sīla ‘conduct’
Set III Sanskrit Prakrit 20. kāśmīra kamhīra ‘Kashmir’ 21. grīṣ̣ma grimha ‘summer’ 22. vismaya vimhaya ‘surprise’ 23. ūṣman umhā ‘heat’ 24. viṣṇu viṇhu ‘Visnu’ 25. praśna paṇha ‘question’ 26. snāna ṇhāṇa ‘bath’ (Bhat 2001: 6–7, 32, 83)
Write out the changes, and identify (by name) the kind of changes where possible.
13 (1) Consonant assimilation: C1 > C2[unaspirated] /C2[non-nasal] __ (C1 not = h) By (1), in consonant clusters, the second consonant assimilates completely to match the first consonant except that if the second consonant is a voiced aspirate, the first consonant assimilates in all other features, but does not become aspirated. (1.-11., 14.- 16.) (2) Merger: š > s (š, s > s) (13., 15., 17., 19. (and 1., 12., 14., 16., 18.)) (3) Metathesis: Sibilant Nasal > Nasal Sibilant (20.-26.) (Sibilant: s, š, ṣ) (4) Sibilant > h: Sibilant > h /Nasal __ (20.-26.) A sibilant (s, š, ṣ) becomes h when following a nasal (many cases with original sibilants end up with the sibilant in the position after a nasal do to the metathesis of (3), and these become h by (4)). It is possible to formulate a change that combines (3) and (4) in a single change: Sibilant Nasal > Nasal h However, this it is unlikely that the language both metathesized sibilants and nasals and at the same time turned all the sibilants into h in a single change. It is more likely, more phonetically plausible, that the change took place in two steps, as in (3) ane (4). There is no evidence in the data to determine whether (2) Merger took place before or after (3) Metathesis. Of course, it is also possible that the changes took place with first (4) turning all sibilants to h before nasals and then later (3) took place to move the new h to follow the nasals. This would require a restatement of (3) to replace the sibilants of the change as stated here with h instead.
Exercise 2.4 Sound change – Proto-Slavic to Russian
What sound changes that have taken place in Russian since Proto-Slavic times are illustrated in the following data? Write rules to account for the palatalization of consonants, the change in the stem vowels, loss of vowels, and change in voicing of consonants. Do not attempt to write sound change rules for the changes in the consonant clusters (bl, tl, dl) in examples 1, 2, and 3. More than one change has applied to some forms; for these, state the relative chronology of these changes (the order, temporal sequence) in which the different changes took place. (The breve /˘/ over vowels means ‘short’.) [[NOTE: for all instances of ɪ̆ and ʊ̆ here, the breve marks (the ̆ ) go directly over (above) the ɪ and ʊ, respectively.]]
Proto-Slavic Russian 1. *greblʊ̆ grjop ‘rowed’ 2. *metlʊ̆ m jol ‘swept’ 3. *vedlʊ̆ vjol ‘led’ 4. *nesʊ̆ njos ‘carried ’ 5. *p̆ɪsʊ̆ pjos ‘dog’ 6. *domʊ̆ dom ‘house’ 7. *grobʊ̆ grop ‘coffin’ 8. *nosʊ̆ nos ‘nose’ 9. *rodʊ̆ rot ‘gender’ 10. *volʊ̆ vol ‘bull’
14 11. *dɪ̆nɪ̆ djenj ‘day’ 12. *konɪ̆ konj ‘horse’ 13. *vɪ̆sɪ̆ vjesj ‘all’ (Verb forms in these data = ‘third person masculine past tense’)
NOTE: This exercise is considered difficult to solve, though easy to understand once the changes are discovered. What sound changes that have taken place in Russian since Proto- Slavic times are illustrated in the following data? Write changes to account for the palatalization of consonants, the change in the stem vowels, loss of vowels, and change in voicing of consonants.
(1) Palatalization: C > Cj /__ Front Vowel (e, ɪ̆) (1.-5., 11.-13.) (C > Cj /__ e, ɪ̆) Consonants were palatalized before front vowels (e and ɪ̆ in these data). (2) V-Rounding: Front V > Round /__ Cʊ̆ (1.-5., 1l., 13.) (e, ɪ̆ > o, u /__Cʊ̆) Front vowels were rounded and turned into corresponding back vowels when followed in the next syllable by ʊ̆. (3) Apocope: ɪ̆, ʊ̆ > Ø /__ # (1.-13.) Final lax vowels were deleted. (4) Lax V-lowering: Lax V > [-high] (11., 13.) (ɪ̆ > e, ʊ̆ > o) Remaining lax vowels were lowered (ɪ̆ > e, ʊ̆ > o). (5) Final-devoicing: stop > voiceless /__ # (7., 9. (1.)) Stops became voiceless word-finally. For (2), an alternative would be to have in place of ʊ̆ in the environment just Round Vowel, which would be true of these data since ʊ̆ is the only Round Vowel that appears in this context. Another alternative would be instead of front vowels becoming round, to state the change as front vowels becoming o (Front V > o /__ Cʊ̆), since ultimately all front vowels that survive in this context end up as o. In the version that says these vowels are rounded, ɪ̆ goes j j j j through an intermediate stage to ʊ̆ (p ɪ̆sʊ̆ > p ʊ̆sʊ̆) and then is lowered to o by (3) (p ɪ̆sʊ̆ > p ʊ̆sʊ̆ j j > p osʊ̆), and then finally > p os by (4)). If (2) is stated as Front V > o in this environment, then j j p ɪ̆s > p osʊ̆ directly by (2), and as a consequence (4) would then not apply to forms such as this. For (3) Apocope, several alternatives also fit the data: (3a) V > Ø /__ # (all final vowels are deleted, true of these data, but not really true if other data were brought into the picture.) (3b) High V > Ø /__ # (final high vowels are lost; the change, stated in this way, predicts any non-high vowel that might appear in this environment would be deleted, though in these data the only high vowels at appears in that context are ɪ̆ and ʊ̆.) (3c) Lax V > Ø /__ # (the weak or lax ɪ̆ and ʊ̆ are deleted finally, but if any non-lax vowels should appear in this environment, the change stated in this way predicts they would not be deleted.) Relative chronology: state the relative chronology of these changes (the order, temporal sequence) in which the different changes took place. 4. carried 5. dog 7. coffin 11. day 12. horse 13. all Proto-Slavic *nesʊ̆ *pɪ̆sʊ̆ *grobʊ̆ *dɪ̆nɪ̆ *konɪ̆ *vɪ̆sɪ̆ j j j j j j j (1) Palatalization: n esʊ̆ p ɪ̆sʊ̆ -- d ɪ̆n ɪ̆ kon ɪ̆ v ɪ̆s ɪ̆ j j (2) V-Rounding: n osʊ̆ p usʊ̆ ------j j j j j j j (3) Apocope: n os p us grob d ɪ̆ n kon v ɪ̆s
15 j j j j j (4) Lax V-lowering: -- p os -- d en -- v es (5) Final-devoicing: -- -- grop ------Result: njos pjos grop djenj konj vjesj In temporal order, (1) Palatalization took place before (2) Vowel Rounding, since if 4. *nesʊ̆ ‘carried’ had first changed to nosʊ̆ by (2) Vowel-Rounding, then (1) Palatalization could not have applied later, since there would no longer have been a front vowel after the consonant to cause it to become palatalized: *nesʊ̆ (2) V-Rounding: nosʊ̆ (1) Palatalization: -- (3) Apocope: nos Result ✘nos Change (1) Palatalization happened before change (3) Apocope, since otherwise vowels that condition palatalization would be gone before palatalization took place and we would not be able to explain why some consonants end up palatalized and others not (as illustrated with 12. *konɪ̆ ‘horse’): *konɪ̆ (3) Apocope: kon (1) Palatalization: -- Result ✘kon Change (5) Final-devoicing must have taken place after (3) Apocope since otherwise the wrong form would result. If in 7. *grobʊ̆ ‘grave’ (5) had taken place first, there would not yet be any final b to change do p, so that when (3) applied later, the result would be the inaccurate ✘grob: *grobʊ̆ (5) Final devoicing: -- (3) Apocope: grob Result ✘grob
Exercise 2.5 Sound change in dialects of Tulu (Dravidian) The forms in the Sapaliga dialect correspond to those of the oldest stage of the language; therefore, compare the forms in the other dialects to those of Sapaliga and determine what sound changes have taken place in each of the other dialects of Tulu. Write out and list the sound changes for each dialect, and identify (name) the kind of change involved in each instance, wherever this is possible. Do you imagine that some of the dialects went through more than one change in intermediate stages to arrive at some of the individual sounds they now have? If so, what might the intermediate stages have been? NOTE:
Sapaliga Holeya Setti Jain 1 Jain 2 1. tare care sare hare are ‘wear off’ 2. tali cali sali hali ali ‘sprinkle’ 3 tavdu cavdu savdu havdu avdu ‘bran’ 4. tōjɨ cōjɨ sōjɨ hōjɨ ōjɨ ‘appear’
16 5. tinɨ cinɨ sinɨ hinɨ inɨ ‘eat’ 6. tudɛ cudɛ sudɛ hudɛ – ‘river’ 7. tōdu cōdu sōdu hōdu ōdu ‘stream’ 8. tanɛ canɛ sanɛ hanɛ anɛ ‘conceiving’ (of cattle) 9. tappu cappu sappu happu appu ‘leaf’ 10. tay cay say hay ay ‘die’ 11. tavtɛ cavtɛ savtɛ havtɛ avtɛ ‘cucumber’ 12. tuttu cuttu suttu huttu uttu ‘wear’ 13. tumbu cimbu sumbu humbu umbu ‘carry on head’ 14. tū cū sū hū ū ‘see’ (Bhat 2001: 51)
Holeya changes: t > c (that is, t > č) Note in 13. tumbu > cimbu, since there is only one example in the data of this difference, it is possible to speculate about several possible changes that would account for the difference, for example perhaps u > i /__labial CCu.
Setti changes: t > s Perhaps more plausibly through the intermediate stage: t > č > s)
Jain 1 changes: t > h Perhaps more plausibly through the intermediate stages: t > č > s > h)
Jain 2: t > Ø Perhaps more plausibly through the intermediate stages: t > č > s > h > Ø)
Exercise 2.6 Sound change Proto-Indo-European (PIE) to Latin
What sound changes have taken place in the transition from Proto-Indo-European to Latin? Try to formulate the most general, most inclusive statement possible to describe these changes. Concentrate on the sounds at the beginning of roots. (The sounds h1, h2, and h3 refer to the laryngeals reconstructed for Proto-Indo-European; do not be concerned with their phonetic value but just treat them as consonants of a general sort. The laryngeals h2 and h3 are thought to represent some sort of fricatives produced in the back of the mouth, with h1 frequently equated with glottal stop.) . Proto-Indo-European Latin 1. *pórk̂os ‘piglet’ porcus /porkus/ ‘young pig’ 2. *pótis ‘capable’ potis ‘capable’
3. *bhébhrus ‘beaver’ fiber ‘beaver’ h 4. *b réh2tēr ‘brother’ frāter ‘brother’ 5. *bhér-o ‘carry’ ferō ‘I carry’
6. *tauros ‘bull’ taurus ‘bull’
17 7. *tréyes ‘three’ trēs ‘three’
8. *dl̥kus ‘sweet’ dulcis /dulkis/ ‘sweet’ 9. *dóh3nom ‘gift’ dōnum ‘gift’
h 10. *d eh1lus ‘nourishing’ fēlīx /fe:li:ks/ ‘fruitful’ h 11. *d úh2mos ‘smoke; fūmus ‘smoke’ h 12. *d eh1- ‘to place, set’ faciō /fakio:/ ‘I do, make’
13. *k̂m̥tóm ‘hundred’ centum /kentum/ ‘hundred’ 14. *k̂(u)wōn ‘dog’ canis /kanis/ ‘dog’
15. *ĝénu ‘jaw’ gena ‘cheek’ 16. *ĝus-tu- ‘taste’ gustus ‘taste’
17. *ĝhans- ‘goose’ hanser (later ānser) ‘goose’ 18. *ĝhais-e- ‘to adhere’ haerē- ‘to stick, cling’
19. *kápr̥̥ ‘penis’ caper /kaper/ ‘goat’ 20. *kom ‘near, by, with’ cum /kum/ ‘with’
21. *genh1os ‘race, kind’ genus ‘race, kind’ 22. *gl̥h1is ‘mouse’ glīs ‘dormouse’
23. *ghaidos ‘goat’ haedus ‘young goat’ 24. *ghóstis ‘stranger, guest’ hostis ‘host’
25. *kwis ‘who’ quis /kwis/ ‘who’ 26. *kwétwor- ‘four’ quattuor /kwattuor/ ‘four’
27. *gwemyo ‘come’ veniō /wenio:/ ‘I come’ w 28. *g ih3wos ‘living’ vīvus /wiwus/ ‘alive’
29. *gwhermós ‘warm’ formus ‘warm’ 30. *gwhén- ‘repel’ (dē-)fen(-dō) ‘I repel’ 31. *gwhér- ‘wild beast’ ferus ‘wild’ [Note that the diacritic small circle goes directly beneath the letter and the small wedge goes directly above the letter it is associated with in these examples.]
What sound changes have taken place in the transition from Proto-Indo-European to Latin? (1) Non-labialized velar and palatal aspirates > h (*gh, *ĝh > h) (as in 17.-18., 23.-24.) (2) Voiced aspirates > f (*bh, *dh, *gwh > f) (as in 3.-5., 10.-12., 29.-31.) (3) Palatal stops > velar (*k̂ > k; *ĝ > g) (1., 13.-16.) The remaining palatal stops become velars; *ĝh had become h by (1) and so does not enter this change. (4) *gw > w (in 27.-28.)
18 Other stops remain unchanged (p, t, k, d, g, kw). (As in 1.-2., 4., 6.-9., 13., 16., 19.-26.) [NOTE: Proto-Indo-European appears not to have had *b, though there is some difference of opinion over this.] Relative chronology: (1) (*gh, *ĝh > h) took place before (2) (voiced aspirates > f), since if (2) had taken place first, then *gh and *ĝh would first have become f as the other voiced aspirates did, with no subsequent means of distinguishing the f’s that came from *gh or *ĝh from the f’s which came from the other sounds, so no way for these words to go *gh, *ĝh > f > h, without also changing the f not originally from *gh or *ĝh that occurs in other words. Changes (2), (3), and (4) could have taken place in any sequence, and their temporal order cannot be determined from the form of the changes themselves. For example, the palatal stops in (3) could have become velars either before or after (2) in which voiced aspirates became f. The change in (4) (*gw > w) could have taken place before any of the other changes or after all of them, or in any other order, since neither the original *gw nor the resulting w interacts with the sounds in any of these other sound changes.
Exercise 2.7 Portuguese sound changes
Make the most general statements you can to account for the sound changes which took place intervocalically in the transition from late Latin to modern Portuguese, based on the following examples. (Do not attempt to explain changes in vowels that you see.)
1. lupu lobo ‘wolf’ 2. sapore sabor ‘flavor’
3. mutu mudo /mudu/ ‘dumb, silent’ 4. latu lado /ladu/ ‘side’ 5. caritate caridade ‘charity’
6. pacare /pakare/ pagar ‘to pay’ 7. focu /foku/ fogo /fogu/ ‘fire’
8. nebula névoa ‘mist’ 9. debere dever ‘to owe’ 10. caballu cavalo /kavalu/ ‘horse’
11. gradu grau ‘degree’ 12. nuda nua ‘nude’ (feminine)
13. regale real ‘royal’ 14. cogitare ‘to think’ cuidar ‘to take care’
15. palu pau ‘stick’ 16. filu fio /fiu/ ‘thread’ 17. salute saúde ‘health’
19 18. luna lua ‘moon’ 19. corona coroa ‘crown’ 20. moneta moeda ‘coin’
Sound changes: (1) Lenition and loss: voiced stops were lenited between vowels: (1a) b > v /V__V (1b) non-labial voiced stop > Ø /V__V (d, g > Ø) (2) Voicing: voiceless stop > voiced /V__V [[Note that it is not actually necessary to specify that the stops entering this change were voiceless, since if we say simply that stop > voiced, the result is the same, with all intervocalic stops ending up voiced, though any that happened already to be voiced would simply remain voiced, undergoing the change vacuously, so to speak.]] (3) Deletion: single alveolar non-obstruent > Ø /V__V (l, n, r > Ø between vowels) An alveolar sonorant (liquid or nasal) was deleted between vowels. Relative chronology: (1) lenition of voiced stops took place before (2) voicing of intervocalic voiceless stops; if the changes had occurred in the other termporal order, (2) before (1), then the voiceless stops p, t, k would have become voiced (b, d, g) first and then these new voiced stops from former voiceless stops via (2) would also have gone on to be lenited, b to v by (1a) and d and g to Ø by (1b), so, for example 1. sapore > sabor flavour’ by (2), and then sabor > ✘savor by (1a); for 4. latu > ladu ‘side’ by (2) first and then ladu > ✘lau by (1b). Change (3) Deletion, however, could have happened at any time relative to the other changes, before or after (1) Lenition, or before or after (2) Voicing, or before both (1) and (2), since the alveolar non-obstruents of (3) do not interact with the stops of (1) and (2).
Exercise 2.8 Some Greek sound changes What happened to the former labiovelar sounds and *w in Attic Greek? Formulate the most precise, most inclusive statement of the sound change(s) from Proto-Indo-European (PIE) to Attic Greek that you can based on the following data.
PIE Attic Greek 1. *kwis tis ‘who’ 2. *kwe te ‘and’ 3. *kwetwóres tettares ‘four’ 4. *pénkwe pente ‘five’
5. *kwóti- ‘how much, pósis ‘how much, how many’ how many’ 6. *kwóteros ‘which of two’ póteros ‘which’ 7. *leikwo- leipō ‘I leave’ 8. *yēkwr̥ hēpar ‘liver’
9. *gwelbhus delphús ‘womb’ 10. *gwous bous ‘cow’ 11. *gwm̥ti- basis ‘going’
20 12. *gwm̥ye- ‘come’ bainō ‘I walk, come’ 13. *gwabh- bap-tō [p < ph] ‘I dip in’
14. *gwhaidrós ‘bright, shining’ phaidrós ‘beaming, cheerful’ 15. *gwhren- ‘think’ phrenéō ‘I think’ 16. *snigwh-s nipha ‘snow’
17. *gwhermós thermós ‘warm’ 18. *gwhel- ‘wish, want’ thélō ‘I want, wish’ 19. *gwhen- ‘strike’ theínō ‘I strike’ 20. *gwhónos ‘stiking down’ phónos ‘murder’
21. *woik̂- ‘clan’ oikía ‘household’ 22. *wóghos ‘carrier’ okhos ‘wagon’ 23. *h2ówis ois ‘sheep’ 24. *dhewo- ‘run’ theō ‘I run’ [Note that the diacritic small circle goes directly beneath the letter and the small wedge goes directly above the letter it is associated with in these examples.]
Sound changes: (1) Labiovelar > alveolar before i or e (*kw, *gw, *gwh > t, d, th, respectively, /__ i, e) (As in 1.-4., 9., 17.-19.) (2) Labiovelar > labial (*kw, *gw, *gwh > p, b, ph, respectively) elsewhere (when not before i or e) (As in 5.-8., 10.-16., 20.) (3) w > Ø /__o (As in 21.-24.)
Relative chronology: Depending on how the sound changes are formulated, there may be no instances of relative chronology, where any of these changes could have taken place in any temporal order with respect to any of the other changes. For example, (3) w-loss could have happened before either (1) and (2) or after one or the other or both of them, since w does not interact with the stop consonants specified in (1) and (2). As stated, (1) [labiovelar > alveolar before i or e] of necessity would have taken place before (2), since otherwise all labiovelars would have ended up as labialsby (2), leaving no labiovelars before i or e to change by (1). Thus if the sequence had been (2) before (1), then, for example, 1. *kwis > ✘pis ‘who’ by (2), leaving no kw to change to t before the i by (1) in this form. However, the changes could be stated differently, with (2) spelled out as taking place specifically before a, o, u and consonants, which would then allow the possibility for the changes to take place in the order (1) before (2) [as restated] or as (2) before (1), since now the two changes specify different, non-overlapping contexts (and can, in fact be combined into a single change by the notation available to us), so that no wrong forms result if the changes applied in the either sequence.
Exercise 2.9 Sound change – Balto-Finnic
21 Determine what sound changes affecting the vowels have taken place in Finnish and Estonian. Write the rules which specify these changes and under what conditions they took place. Identify (name) the changes, where possible. NOTE: ä = [æ], ö = [ø], ü = [y], õ = [ɨ]. Double vowels (for example aa, oo, and so on) are long vowels. Orthographic
Proto-Balto- Finnish Estonian gloss Finnic 1. *maa maa maa ‘land’ 2. *noori nuori noor ‘young’ 3. *koori kuori koor ‘bark, peel’ 4. *hooli huoli hool ‘care, worry’ 5. *jooni juoni joon ‘line, direction’ 6. *leemi liemi leem ‘broth’ 7. *mees mies mees ‘man’ 8. *meeli mieli meel ‘mind’ 9. *keeli kieli keel ‘tongue, language’ 10. *reemu riemu rõõm [rɨ:m] ‘joy’ 11. *meekka miekka mõõk [mɨ:k:] ‘sword’ 12. *peena piena põõn [pɨ:n] ‘slat, rail, cross-piece’ 13. *veeras vieras võõras [vɨ:ras] ‘foreign’ 14. *luu luu luu ‘bone’ 15. *hiiri hiiri hiir ‘mouse’ 16. *kyynärä kyynärä küünar ‘ell (measure)’ 17. *töö työ töö ‘work’ 18. *möö- myö- möö- ‘along, by’ 19. *kala kala kala ‘fish’ 20. *lapa lapa laba [lapa] ‘blade’ 21. *kylä kylä küla ‘village’ 22. *ikä ikä iga [ika] ‘age’ 23. *isä isä isa ‘father’ 24. *joki joki jõgi [jɨki] ‘river’ 25. *kivi kivi kivi ‘stone’ 26. *lumi lumi lumi ‘snow’ 27. *läpi läpi läbi [læpi] ‘through, hole’ 28. *suku suku sugu [suku] ‘family’ 29. *ilma ilma ilm ‘world’ ‘weather, world’ 30. *jalka jalka jalg [jalk] ‘foot, leg’ 31. *kalma kalma kalm ‘grave (mound)’ 32. *nälkä nälkä nälg [nælk] ‘hunger’ 33. *härkä härkä härg [hærk] ‘ox, bull’ 34. *silmä silmä silm ‘eye’ 35. *marja marja mari ‘berry’ 36. *karja karja kari ‘cattle’
22 37. *orja orja ori ‘slave’ 38. *lintu lintu lind [lint] ‘bird’ 39. *hullu hullu hull ‘crazy’ 40. *mänty mänty mänd [mænjtj] ‘pine’ 41. *synty synty sünd [synjtj] ‘birth’ 42. *hanki hanki hang [haŋk] ‘crust of snow’ 43. *kurki kurki kurg [kurk] ‘crane’ 44. *nahka nahka nahk ‘leather’ 45. *lehmä lehmä lehm ‘cow’ 46. *lehti lehti leht ‘leaf, sheet’ 47. *hauta hauta haud [haut] ‘grave’ 48. *lauta lauta laud [laut] ‘board’ 49. *lava lava lava ‘platform, frame’ 50. *haava haava haav ‘wound’ 51. *hinta hinta hind [hint] ‘price’ 52. *into into ind [int] ‘passion’ 53. *halko halko halg [halk] ‘piece/block of wood’ 54. *kylmä kylmä külm ‘cold’ 55. *hullu hullu hull ‘crazy’ 56. *hiki hiki higi [hiki] ‘sweat’ 57. *kylki kylki külg [kylk] ‘side’ 58. *kirppu kirppu kirp [kirp:] ‘flea’ 59. *verkko verkko võrk [vɨrk:] ‘net’ 60. *onsi onsi õõs [ɨ:s] ‘a hollow place’ 61. *kansi kansi kaas ‘cover’ 62. *kynsi kynsi küüs ‘fingernail, claw’ 63. *mesi mesi mesi ‘honey’ 64. *kuusi kuusi kuus ‘six’ 65. *kusi kusi kusi ‘urine’ 66. *mato mato ‘worm’ madu [matu] ‘snake’ 67. *elo elo elu ‘life/building’ 68. *hako hako hagu [haku] ‘evergreen sprig, brushwood’ 69. *ilo ilo ‘joy’ ilu ‘beauty’ 70. *himo himo himu ‘lust, desire’ 71. *iho iho ihu ‘skin, hide’ 72. *vesa vesa võsa [vɨsa] ‘sprout, brush, weed’ 73. *helma helma hõlm [hɨlm] ‘skirt, frock’ 74. *terva terva tõrv [tɨrv] ‘tar’ 75. *velka velka võlg [vɨlk] ‘debt’ 76. *perna perna põrn [pɨrn] ‘spleen’ 77. *leuka leuka lõug [lɨuk] ‘jaw, chin’ 78. *tosi tosi tõsi [tɨsi] ‘true’ 79. *solki solki sõlg [sɨlk] ‘buckle, brooch’ 80. *sormi sormi sõrm [sɨrm] ‘finger’ 81. *pohja pohja põhi [pɨhi] ‘bottom, base’ 82. *poski poski põsk [pɨsk] ‘cheek’
23 83. *korpi korpi kõrb [kɨrp] ‘dark woods, wilderness’ 84. *metsä metsä mets ‘woods’ 85. *leppä leppä lepp [lep:] ‘alder’
NOTE: the notation Co means any number of consonants from Ø (no consonant at all) to any number or consonants in a consonant cluster; it is essentially a device to say we are concerned mostly with syllables or with vowels, and the number of possible intervening consonants is not relevant to the change.
Sound change in Finnish: (1) Long mid-vowel diphthongization: long mid V > diphthong with high first component (ee > ie, oo > uo, öö > yö) (as in 1.–18., etc.)
Sound changes in Estonian:
(1) e-backing: e > ˆ /C__Co back V (e > ˆ, ee > ˆˆ when non-initial in the word and followed in the next syllable by a back vowel) (as in 10.–13., 59., 71.-76.)
(2) o-fronting and raising: o > ˆ /__Co front V (o > ˆ when followed in the next syllable by a front vowel, where in these data i is the only front vowel in that context) (as in 24., 59., 77.-82.) (3) Palatalization: t > tj /__ y (as in 40.-41.) [Note that y here is the the same as the IPA symbol, indicating a high front rounded vowel here, not a glide (not IPA [j]).] (t is palatalized before high front rounded y.) With more data, this change would be seen to take place before i as well. It might be better formulated as: t > tj /__ i, y (that is, t > tj /__ high front V). (4) Apocope: V > Ø /VVC__#, /VCC(C)__# (A word-final vowel is lost if the preceding syllable is ‘heavy’, that is, if it has a long vowel or has a consonant cluster [including geminate consonants]) (as in 2.-12., 15., 29.-54., 56.-61., 63., 72.-76., 78.-84.) In 16., the final vowel is lost in a three-syllable word, but this is the only example in the data; to account for it, we need to add to the context of the change something equivalent to /V(V)CoVCo__#, which would predict that the final vowel of any three-syllable word would be lost. If you believe the loss would happen in trisyllabic forms only if the first syllable is long, that would be indicated, for example, as: /VVCoVCo__#. Or if you believe the change would take place in three-syllable words when the first syllable is heavy (either with a long vowel or with a consonant coda), then that would be indicated, for example, as: /VVCoVCo__#, /VCC(C)VCo__#. (5) j-vocalization: j > i /___ # (as in 35.-37., 80.) (Word-final j becomes i. A j could only come to be in final position due to the earlier loss of a final vowel by (4), as in orja > orj by (4), then orj > ori by (5). (6) Nasal assimilation: Nasal > [αPlace] /__[-continuant, αPlace]C (as in 38., 40.-42., 51.-52.) Nasals assimilate to the place of articulation of following stops. The cases with nt are in effect already assimilated (as in 38., 51., 52.) and so no change takes place; 42. with ŋk also was already assimilated but this is not reflected in the orthography (
24 Relative chronology. This exercise does not call the relative chronology to be stated; however, for those who may be curious about this: (1) e-backing, (2) o-fronting and raising, and (3) Palatalization took place before (4) Apocope. If (4) Apocope had taken place first, then none of these three changes could have taken place in some forms that did undergo the changes, since (4) would have deleted the final vowel which conditions these three changes. For example, if they had taken place with (4) first, the results would be wrong: 72. helma 79. sormi 41. synty ‘skirt’ ‘finger’ ‘birth’ (4) Apocope: helm sorm synt (1) e-backing: ------(2) o-fronting and raising: ------(3) Palatalization: ------Result: ✘helm ✘sorm ✘synt Expected: hˆlm sˆrm synjtj (syntj by (3), then synjtj by (6)) Changes (1), (2), and (3) do not interact with one another, since the contexts in which they took place or the sounds which changed are different in all three cases. Change (5) j- vocalization must follow (4) Apocope, since otherwise, the j would not be in final position and thus would not fit the environment where (5) took place, giving the wrong result: 37. orja ‘slave’ (5) j-vocalization: -- (4) Apocope: orj Result: ✘orj Expected: ori In short, the temporal sequence had to be: First (1) e-backing, (2) o-fronting and raising, (3) Palatalization (these three changes could have taken place in any order) Second (4) Apocope Third (5) j-vocalization The changes (6) Nasal assimilation, (7) Nasal loss and compensatory lengthening, and (8) Final o-raising do not interact with one other or with the other changes, and thus they could have taken place in any temporal order.
25
EXERCISES FOR CHAPTER 3 BORROWING
Exercise 3.1
Find ten examples of loanwords (not already mentioned in this chapter) into any language you like, including English. You can consult dictionaries which give historical sources of lexical items or books on the history of particular languages, if you wish. Try to identify the form and meaning of the word in the donor language.
There are many books and websites that identify loanwords in many different languages, which can be consulted.
Exercise 3.2 Twentieth-century loans into English
In the history of English, relatively few words were borrowed during the twentieth century when seen in comparison with the large number of loans from earlier times. Still, many did come into the language; here are a few of them. Look up twenty of these (or more if you like) either in a good dictionary of English which indicates the sources from which words come or in a dictionary of the language from which they were borrowed. Try to determine the original meaning and form in the donor language and note any changes (in meaning or form) that the word has undergone as it was adopted into English. The original meanings of many of these may surprise you.
Afrikaans: apartheid Chinese: chow mein, kung fu Czech: robot French: avant-garde, boutique, camouflage, chassis, cinema, discotheque, fuselage, garage, limousine, sabotage German: angst, blitz, ersatz, flak, Nazi, snorkel, strafe, wienerschnitzel Hawai’ian: aloha, lei, ukulele Hebrew: kibbutz Italian: fascism, partisan, pasta, pizza Japanese: bonsai, kamikaze, karaoke, karate, origami Russian: bolshevik, cosmonaut, glasnost, intelligentsia, perestroika, sputnik Spanish: aficionado, macho, marijuana, paella, tango Swedish (or Scandinavian generally): moped, ombudsman, slalom, smorgasbord Yiddish: schmaltz, schlock, klutz
Afrikaans: apartheid Form: apartheid, composed of apart ‘separate’ + -heid ‘-hood’. Meaning: ‘separateness’ (a South African political term to avoid saying ‘segregation’). Chinese:
26 chow mein The form and meaning are an American creation, ‘sir-fired noodles’: from Taishan Chinese (Guangdong) chāu-mèing ‘fried flour’; based on chow-chow ‘food’ in Chinese English Pidgin, and mein (mèing, mihn) ‘wheat flour, wheat flour noodles’. kung fu Form: kung fu (gongfu) Meaning: ‘effort, energy’, ‘skill, accomplishment’; Chinese martial arts.
Czech: robot From robotnik ‘slave’ (based robotitit ‘work, drudge’), taken from the English 1923 translation of the play R.U.R. (Rossum's Universal Robots), by Karel Capek.
French: avant-garde Form avant-garde Meaning: ‘advance guard’, literally ‘before-guard’. boutique Form: boutique, ultimately from Greek apothēkē ‘repository; granary’; related to other loans in English from the same Greek original word: apothecary, bodega. Meaning: ‘small shop’ (later ‘fashion shop’). camouflage From: French camoufler, Parisian slang, ‘to disguise’, from Italian camuffare ‘to disguise’. chassis From French châssis ‘frame’ (derived from chasse ‘case, box’). cinema From French cinéma, shortened from cinématographe, coined by Lumiere brothers, who invented it based on Greek kinema ‘movement’ (compare English ‘movie’). discotheque From French discothèque, borrowed from Italian discoteca ‘record collection, record library’, coined on the basis of disco ‘phonograph record’ and -teca taken from biblioteca ‘library’. Meaning: ‘nightclub with recorded music for dancing’, ‘record library’. fuselage From French fuselage, from fuselé ‘spindle-shaped, tapered’ (compare Old French fus ‘spindle’. Called this because of its spindle-like shape. garage
27 From French garage ‘vehicle shelter’, originally ‘a place for storing things’, derived from garer ‘to shelter, park’ (ultimately from a Germanic loanword in French). limousine From French limousine, and adjective derived from Limousin, a region of central France. The automobile meaning comes from the car’s similar appearance to a type of hood worn by the inhabitants of Limousin province. Meaning from which modern senses come: enclosed automobile with open driver's seat. sabotage From French sabotage, derived from saboter ‘to sabotage, bungle’, which in turn is from sabot ‘wooden shoe’. The meaning of ‘deliberate and malicious destroying of property’ is said to come from ‘labor disputes’.
German: angst From German Angst ‘anxiety, remorse, neurotic fear’, made popular in English by the translation of Sigmund Freud’s work. blitz From German Blitz ‘lightening’, for a ‘sudden intensive attack’; in English taken from Blitzkrieg (Blitz ‘lightening’ + Krieg ‘war’). ersatz From German Ersatz ‘substitute, replacement’. In English meaning: ‘substituting for, and typically inferior in quality’. flak From German Flak, which comes from Fliegerabwehrkanone ‘anti-aircraft gun’ (composed of Flieger ‘flier, pilot’ + Abwehr ‘defense’ + Kanone ‘canon’). In English it had the sense of ‘anti-aircraft fire, which came to be extended metaphorically to ‘criticism’.
Nazi From German Nazi, an abbreviation of Nationalsozialist ‘national socialist’, from Nationalsozialistische Deutsche Arbeiterpartei ‘National Socialist German Workers’ Party’, which Adolf Hitler led. snorkel From German Schnorchel, slang for ‘nose, snout’. strafe From German strafen ‘to punish’, used in English for ‘punish, attack, especially attack from low-flying aircraft. wienerschnitzel
28 From German, Wiener ‘someone or something from Wien ‘Vienna’ + Schnitzel ‘cutlet, a slice’, derived from Schnitz ‘a cut’ (see schnitzen ‘to carve’) + -el ‘diminutive’.
Hawai‘ian: aloha From Hawai‘ian aloha ‘love, affection, greeting’. lei From Hawai‘ian lei ‘garland, wreath, necklace of flowers’. ukulele From Hawai‘ian ‘ukulele, from ‘uku flea, louse’ + lele ‘to fly, jump, leap’. So called from the rapid motion of the fingers in playing it. It developed from a Portuguese instrument introduced to the islands c.1879.
Hebrew: kibbutz From Hebrew qibbus ‘gathering’. In English: ‘Israeli collective settlement’.
Italian: fascism Derived from fascist, from Italian partito nazionale fascista, Benito Mussolini’s anti- communist political movement, from Italian fascio ‘bundle, association, group’. pasta From Italian pasta ‘dough, paste, pasta’, itself from late Latin pasta ‘dough, cake, paste’, borrowed from Greek pasta ‘barley porridge’. pizza From Italian pizza, originally ‘cake, pie’. Its origin is uncertain; it may be connected with Greek pita ‘pita, flatbread’.
Japanese: bonsai From Japanese bon ‘basin, pot’ + sai ‘to plant’. kamikaze From Japanese kami ‘divine, god, providence’ + kaze ‘wind’. This was the name given a typhoon in folklore which saved Japan from Mongol invaders by destroying Kublai Khan’s fleet; it was applied to ‘suicide fliers’ in World War II. Meaning: ‘divine wind’. karaoke From Japanese kara ‘empty’ + oke ‘orchestra’, a shortened form of okesutora, borrowed from English orchestra.
29 karate From Japanese kara ‘empty, bare’ + te ‘hand’. origami From Japanese origami, composed of ori ‘fold’ + kami ‘paper’.
Russian: bolshevik From Russian bol'shinstvo ‘majority, based on bol'shiy ‘greater’, from bol'shoy ‘big, great’ (as in the Bolshoi Ballet). The majority faction at the Second Congress of the Marxist Russian Social Democratic Workers’ Party was given the name Bol'sheviki, the plural form, in 1903. It subsequently became the Russian Communist Party. In English it came to mean also ‘an extreme radical’. cosmonaut From Russian kosmonavt, ultimately from Greek kosmos ‘universe’ + nautes ‘sailor (English naval and nautical have the same root). intelligentsia From Russian intelligyentsiya ‘the intellectual class’, based on Latin intelligentia. sputnik From Russian sputnik ‘satellite’ (first launched Oct. 4, 1957), from s- ‘with, together’ + put ‘way, path, journey’ + -nik ‘agent suffix’, literally ‘fellow traveler, traveling companion’. soviet From Russian sovyet ‘council’. gulag (also GULAG) From Russian acronym of Glavnoe Upravlenie Ispravitel'no-Trudovykh Lagerei ‘Chief Administration of Corrective Labor Camps’. In English: a political labor camp.
Spanish: aficionado From Spanish aficionado ‘amateur’, specifically a ‘fan of bullfighting’, based on afición ‘affection’. macho From Spanish macho ‘male animal’, use as an adjective meaning ‘masculine, virile’ (from Latin masculus ‘masculine’ by regular sound changes in Spanish). burrito From Spanish burro ‘donkey, burro’ + -ito ‘diminutive’. The name burrito may come from the appearance of the rolled up tortilla with meat and beans inside, somewhat
30 similar to a donkey’s ear or perhaps to the bedrolls and packs that donkeys carried. nacho From Spanish; the food is attributed to a restaurant cook, Ignacio Anaya, who is said to have invented the dish in the border town of Piedras Negras. Nacho is the nickname for men named Ignacio. cilantro From Spanish cilantro ‘cilantro, coriander’, a variant of Spanish culantro, from Latin coriandrum ‘coriander’ (from which English gets coriander).
Swedish and Norwegian: moped From Swedish moped, a blending from (trampcykel med) mo(tor och) ped(aler) ‘bicycle with motor and pedals’. ombudsman From Swedish ombudsman ‘commissioner, agent’, composed of ombud ‘commission’ (with om- ‘around’ + bud ‘bid, command’) + man ‘man’. Meaning: A government official, especially in Scandinavian countries, who investigates citizens’ complaints against the government or its functionaries. slalom From Norwegian slalåm ‘a race on skis’, from slad ‘sloping + låm ‘path’. smorgasbord From Swedish smörgåsbord ‘open sandwich table’, composed smörgås ‘bread and butter’ (from smör ‘butter’ [related to English smear] + gås ‘lump of butter’ [historically from ‘goose’ (related to English goose) + bord ‘table’ (compare English board). In English: ‘a ski race down a winding course marked by flags or poles’.
Yiddish: klutz From Yiddish klots ‘clumsy person, blockhead’. (Compare German klotz ‘boor, clod’, literally ‘wooden block’.)
maven (mavin) From Yiddish meyvn, from Hebrew mevin (mebhin) ‘one who understands’. Meaning: a trusted expert in a particular field. putz From Yiddish puts ‘ornament’. In earlier slang it had the sense of ‘penis’ and then ‘fool, idiot’. Compare schmuck (shmuck) ‘obnoxious, contemptible or detestable person, stupid
31 or foolish person’, from Yiddish schmock ‘jewel, penis’; cf. German schmuck ‘jewel, adornment’. (From the meaning ‘penis’, as in ‘the family jewels’.) schmaltz From Yiddish shmalts ‘melted fat’, extended to mean ‘excessive sentimentality, especially in music’. tush From Yiddish, a shortened form of tukhus (tochus) from Yiddish tokhes ‘behind, backside’, from Hebrew tahat ‘beneath’. Meaning: ‘backside, buttocks’.
Exercise 3.3 Māori and English loanwords
(1) Based on the criteria for establishing loanwords and the direction of borrowing, determine from the following lists of words which are borrowed into Māori from English and which are borrowed into English from Māori. Note that Māori has the following inventory of sounds: /p, t, k, φ, h, r, m, n, ŋ, r, i, e, a, o, u/. In the traditional orthography, /φ/ (voiceless bilabial fricative) is spelled wh; /ŋ/ is spelled ng. Also, native Māori words permit no consonant clusters, rather only syllables of the shape CV (a single consonant followed by a single vowel). (2) Can you say anything about the pronunciation of the variety of English from which Māori took its English loans? (3) What can you say about the social or cultural nature of the contact between speakers of Māori and English? Can you identify semantic domains (areas of meaning) most susceptible to borrowing in either of the languages? (4) How were words from one language modified to fit the structure of the other?
1. hāhi ‘church’ 2. haina ‘China; sign’ 3. haka ‘haka, Māori dance’ 4. haki ‘flag’ (< Union Jack) 5. hāma ‘hammer’ 6. hānara ‘sandal’ 7. hāngi ‘hangi, oven’ (hole in the ground with wrapped food placed on heated stones in the pit with fire) 8. hānihi ‘harness’ 9. hāpa ‘harp’ 10. hāte ‘shirt’ 11. hēmana ‘chairman’ 12. hereni ‘shilling’ 13. heti ‘shed’ 14. hipi ‘sheep’ 15. hiraka ‘silk’ 16. hiriwa ‘silver’ 17. hoeha ‘saucer’ 18. hohipere ‘hospital’
32 19. hopa ‘job’ 20. hōro ‘hall’ 21. hū ‘shoe’ 22. hui ‘meeting for discussion’ 23. huka ‘sugar’ 24. hūka ‘hook’ 25. hupa ‘soup’ 26. hūri ‘jury’ 27. iāri ‘yard’ 28. ihipa ‘Egypt’ 29. ingarangi ‘England’ 30. ingarihi ‘English’ 31. inihi ‘inch’ 32. iota ‘yacht’ 33. iwi ‘iwi, Māori tribe’ 34. kāka ‘cork’ 35. kānara ‘colonel’ 36. kapa ‘copper, penny’ 37. kāpara ‘corporal’ 38. kāpata ‘cupboard’ 39. kara ‘collar’ 40. karaehe ‘grass; glassware, tumbler; class’ 41. karāhi ‘glass’ 42. karahipi ‘scholarship’ 43. karaka ‘clock; clerk’ 44. karauna ‘crown’ 45. kāreti ‘college; carrot; carriage’ 46. kāta ‘cart’ 47. kātaroera ‘castor oil’ 48. kātipa ‘constable’ 49. kaumātua ‘kaumatua, Māori elder’ 50. kauri ‘kauri tree’ 51. kāwana ‘governor’ 52. kea ‘kea’ (mountain parrot) 53. kihi ‘kiss’ 54. kirihimete ‘Christmas’ 55. kiwi ‘kiwi bird’ 56. kōmihana ‘commission’ 57. kōti ‘court (of law); goat’ 58. kuihipere ‘gooseberry’ 59. kūmara ‘kumara, sweet potato’ 60. kura ‘school’ 61. māhi ‘mast’ 62. mana ‘mana, influence, prestige’ 63. māori ‘Māori, native people’ (in Māori māori means ‘clear, ordinary, native New Zealander’)
33 64. marae ‘marae, enclosed meeting area’ 65. marahihi ‘molasses’ 66. moa ‘moa’ (very large extinct flightless bird) 67. mokopuna ‘mokopuna, grandchild’ 68. motokā ‘car, automobile’ (< motor car) 69. nēhi ‘nurse’ 70. ngaio ‘ngaio, coastal shrub’ 71. ōkiha ‘ox’ 72. ōriwa ‘olive’ 73. otimira ‘oatmeal’ 74. pā ‘pa, stockaded village’ 75. pahi ‘bus’ 76. paihikara ‘bicycle’ 77. paitini ‘poison’ 78. pāka ‘box’ 79. pākehā ‘pakeha, European, non-Māori’ 80. pāmu ‘farm’ 81. pāoka ‘fork’ 82. parakuihi ‘breakfast’ 83. parama ‘plumber’ 84. pāua ‘paua, abalone shell’ 85. pāuna ‘pound’ 86. perakēhi ‘pillowcase’ 87. pereti ‘plate’ 88. pī ‘bee’ 89. pirihi ‘priest’ 90. pirihimana ‘police(man)’ 91. piriniha ‘prince’ 92. piriti ‘bridge’ 93. pōkiha ‘fox’ 94. pōro ‘ball’ 95. pukapuka ‘book’ 96. pūkeko ‘pukeko, swamp hen’ 97. pune ‘spoon’ 98. purū ‘blue’ 99. pūru ‘bull’ 100. rare ‘lolly, sweets’ 101. rata ‘doctor’ 102. reme ‘lamb’ 103. rērewē ‘railroad, railway’ 104. rēwera ‘devil’ 105. rīhi ‘dish; lease’ 106. rimu ‘rimu, red pine’ 107. rōre ‘lord’ (title) 108. rori ‘road’ 109. takahē ‘takahe, bird species’ (Notoris mantelli)
34 110. tana ‘ton’ 111. tangi ‘tangi, Māori mourning or lamentation’ (associated with funerals) 112. tāone ‘town’ 113. taonga ‘taonga, heritage, Māori treasure, possessions’ 114. tāra ‘dollar’ 115. taraiki ‘strike’ 116. tauiwi ‘tauiwi, non-Māori’ 117. tēpu ‘table’ 118. tiā ‘jar’ 119. tiaka ‘jug’ 120. tiamana ‘chairman; German’ (cf. hēmana) 121. tiāti ‘judge’ 122. tīhi ‘cheese’ 123. tōtara ‘totara’ (tree species, Podocarpus totara) 124. tui ‘tui, parson bird’ 125. waka ‘waka, canoe’ 126. wātene ‘warden’ 127. weka ‘weka, woodhen’ 128. wētā ‘weta, large insect species’ (Hemideina megacephala) 129. whakapapa ‘whakapapa, genealogy’ 130. whānau ‘whanau, extended family’ (community of close fellows) 131. whatura ‘vulture’ 132. whira ‘violin, fiddle’ 133. whatura ‘vulture’ 134. whira ‘violin, fiddle’ 135. whīra ‘field’ 136. whurū ‘flu’ 137. whurutu ‘fruit’ 138. whutupaoro ‘football’ (rugby) 139. wihara ‘whistle’ 140. wra ‘wheel’ 141. wōro ‘wall’ 142. wuruhi ‘wolf
(1) The following are borrowed from Māori into English: 3. haka ‘haka, Māori men’s dance’ 7. hāngi ‘hangi, oven’ 22. hui ‘meeting for discussion’ 33. iwi ‘iwi, Māori tribe’ 49. kaumātua ‘kaumatua, Māori elder’ 50. kauri ‘kauri tree’ 52. kea ‘kea’ (mountain parrot) 55. kiwi ‘kiwi bird’ 59. kūmara ‘kumara, sweet potato’ 62. mana ‘mana, influence, prestige’ 63. māori ‘Māori, native people’
35 64. marae ‘marae, enclosed meeting area’ 66. moa ‘moa’ (very large extinct flightless bird) 67. mokopuna ‘mokopuna, grandchild’ 70. ngaio ‘ngaio, coastal shrub’ 74. pā ‘pa, stockaded village’ 79. pākehā ‘pakeha, European, non-Māori’ 84. pāua ‘paua, abalone shell’ 96. pūkeko ‘pukeko, swamp hen’ 106. rimu ‘rimu, red pine’ 109. takahē ‘takahe, bird species’ (Notoris mantelli) 111. tangi ‘tangi, Māori mourning or lamentation’ (associated with funerals) 113. taonga ‘taonga, heritage, Māori treasure, possessions’ 116. tauiwi ‘tauiwi, non-Māori’ 123. tōtara ‘totara’ (tree species, Podocarpus totara) 124. tui ‘tui, parson bird’ 125. waka ‘waka, canoe’ 127. weka ‘weka, woodhen’ 128. wētā ‘weta, large insect species’ (Hemideina megacephala) 129. whakapapa ‘whakapapa, genealogy’ 130. whānau ‘whanau, extended family’ (community of close fellows) All the others are loans from English into Māori.
(2) Can you say anything about the pronunciation of the variety of English from which Māori took its English loans? These loans were probably taken from some variety of English with the following properties: [1] Non-rhotic (so-called “r-less” or lacking of non-pre-vocalic “r”), since Māori has /r/ but borrowed such English words without “r”, as in hāma ‘hammer’ (also in 1., 5., 8., 10., 11., 23., 57., 69., 101., 118., etc.) (But see 131. whatura ‘vulture’, which seems to reflect a rhotic pronunciation.) [2] The variety of English may have had final consonant-cluster simplification (for example, mas’ instead of mast, borrowed as 61. māhi, not the expected ✘mahiti), since in cases with consonant clusters, Māori typically broke up the consonants by inserting a vowel (for example, as in: 15., 16., 18., 29., 31., 43., 44., 54., 82., 93., etc.), but did not insert a vowel in these cases, but rather has nothing to represent what appears to be the consonant that was dropped in consonant-cluster simplification (as in 61., 85., 89., 133., etc.). [3] Possibly final “ng” [ŋ] was pronounced as “n’” [n], since Māori had ng [ŋ] and so could have borrowed English words with final “ng” with its ng [ŋ], not with n, as seen in the single example: 12. hereni ‘shilling’, not borrowed as ✘herengi. [4] The English variety may have had l-vocalization, or perhaps a dark “l” (velarized l) sometimes, for which Māori substituted an u as in 117. tēpu ‘table’ (see also 48.). (In varieties of English with vocalized l in this environment, the pronunciation of table is more like [teibo, teibɤ].) However, in most words, Māori substituted r for English final l. [5] The English variety may have had /a/ in words where some other varieties have open “o” ([ɔ]), since they were borrowed in Māori with a, not with o (as in 34., 37., 43., 77., 78.),
36 though this is not true of all forms; some have an o vowel in the borrowing in Māori (see 17.- 20., 71.-72., 81., 93.-94.).
(3) What can you say about the social or cultural nature of the contact between speakers of Māori and English? Can you identify semantic domains (areas of meaning) most susceptible to borrowing in either of the languages?
English borrowed from Māori words for newly encountered plants and animals: 50. kauri ‘kauri tree’ 52. kea ‘kea’ (mountain parrot) 55. kiwi ‘kiwi bird’ 59. kūmara ‘kumara, sweet potato’ 66. moa ‘moa’ (very large extinct flightless bird) 70. ngaio ‘ngaio, coastal shrub’ 84. pāua ‘paua, abalone shell’ 96. pūkeko ‘pukeko, swamp hen’ 106. rimu ‘rimu, red pine’ 109. takahē ‘takahe, bird species’ (Notoris mantelli) 123. tōtara ‘totara’ (tree species, Podocarpus totara) 124. tui ‘tui, parson bird’ 127. weka ‘weka, woodhen’ 128. wētā ‘weta, large insect species’ (Hemideina megacephala)
And Englishe borrowed terms specific to Māori culture and society: 3. haka ‘haka, Māori men’s dance’ 7. hāngi ‘hangi, oven’ 22. hui ‘meeting for discussion’ 33. iwi ‘iwi, Māori tribe’ 49. kaumātua ‘kaumatua, Māori elder’ 62. mana ‘mana, influence, prestige’ 63. māori ‘Māori, native people’ 64. marae ‘marae, enclosed meeting area’ 67. mokopuna ‘mokopuna, grandchild’ 74. pā ‘pa, stockaded village’ 79. pākehā ‘pakeha, European, non-Māori’ 111. tangi ‘tangi, Māori mourning or lamentation’ (associated with funerals) 113. taonga ‘taonga, heritage, Māori treasure, possessions’ 116. tauiwi ‘tauiwi, non-Māori’ 125. waka ‘waka, canoe’ 129. whakapapa ‘whakapapa, genealogy’ 129. whānau ‘whanau, extended family’ (community of close fellows) Māori borrowed much from English, representing plants, animals, foods, and implements of acculturation, aspects of English law, military, religion, society, sports, etc. – representing English influence in many spheres.
(4) How were words from one language modified to fit the structure of the other?
37 English modified Māori words only slightly, since English had all the consonants and vowels Māori has. Vowel length is dropped in English loans, and the initial velar nasal was modified, for example in 70. ngaio ‘coastal shrub’, usually pronounced just /nai.o/ in English. Māori, on the other hand, modified English words considerably to adopt them to the phonological structure of Māori words. [1] Voiced stops were made voiceless (as in 13., 23., 40.-41., 57.-58., 75.-76., 78., 92., 98., etc.). [2] All English sibilants (s, š, č, z, ž, ǰ) were borrowed as /h/ (as in 1.-2., 6., 10.-19., 21., 23., 25.-26., 28., 31., 41.-42., 53.-54., 56., 58., etc.), though sometimes /ti/ is substituted for English /ǰ/ (as in 45., 92., 118.-119., 121.-122.). [3] English consonant clusters are broken up by the insertion of a vowel in the Māori loans (as in 15.-16., 18., 29.-31., 43.-44., 54., 58., etc.). [4] Initial consonant clusters with initial s usually drop the s, as in 60. kura ‘school’, and in115. taraiki ‘strike’, where s is not represented, but tr are, broken up by the insertions of a (for other examples see 42., 97., 137.) [5] Māori adds a vowel at the end of consonant-final words (as in 1., 4., 6., 8.-20., 24.- 32., 34.-35., 38., 40.-47., 53.-54., etc.). [6] The l of English is replaced by r in borrowings into Māori (as in 6., 12., 15.-16., 18., 20., 29.-30., 35., 39., 41.-43., 45., 47., 60., etc.). [7] Māori w usually replaces English /v/ (as in 16., 51., 72., 104). [8] Māori wh replaces English /f/ and sometimes /v/ (as in 131.-136.).
Exercise 3.4 Spanish loanwords
The following is a list of borrowings in Spanish from different languages. What historical and cultural inferences might you suggest about the nature of the contact between speakers of Spanish and each of these other languages based on these? Concentrate on the Germanic and Arabic contacts. Which of the non-Germanic words do you think were further borrowed later from Spanish to English (or from Spanish to French and then on to English)? From Celtic loans, already in Latin (from Gaul), inherited in Spanish: abedul ‘birch tree’, bragas ‘breeches, trousers’, camisa ‘shirt’, carro ‘cart’, cerveza ‘beer’. From Germanic (Swabians in Galicia; Vandals, Alans; Franks – Visigoths entered Spain in AD 412). Loans: eslabón ‘link’, ganar ‘to gain, win, earn’, ganso ‘goose’; bandera ‘flag’, botín ‘booty’, dardo ‘dart’, espiar ‘to spy’, espuela ‘spur’, guardar ‘guard’, guerra ‘war’, guía ‘guide’, hacha ‘axe’, robar ‘to rob’, yelmo ‘helmet’; arpa ‘harp’, banco ‘bench’, barón ‘baron’, blanco ‘white’, brasa ‘live coal’, estaca ‘stake’, falda ‘skirt’, gris ‘grey’, guante ‘glove’, rico ‘rich’, ropa ‘clothing’, sopa ‘soup’, tacaño ‘stingy’, toalla ‘towel’; norte ‘north’, sur ‘south’, este ‘east’, oeste ‘west’; personal names: Anfonso, Elvira, Federico, Fernando, Francisco, Gonzalo, Matilde, Ricardo, Rodrigo; and so on. From Arabic (Moors landed in Spain in AD 711; by 718 they had spread over most of the Peninsula, where they remained until the recapture of Granada in 1492). Loans: Guad- ‘river’ (in place names, for example, Guadalajara ‘river of stones’, Guadarrama ‘river of sand’); alcázar ‘castle’ (corruption of Latin castrum with Arabic article al-), alférez ‘ensign’, alcalde ‘mayor’, atalaya ‘watchtower’, aldea ‘village’, almacén ‘storehouse’, barrio ‘district of city’, adobe (sun-dried brick), albañil ‘mason’, alcoba ‘bedroom’ (alcove), alfarero ‘potter’, bazar
38 ‘bazaar’, alfiler ‘pin’, alfombra ‘rug’, almohada ‘pillow’, ataúd ‘coffin’, aceite ‘oil’, aceituna ‘olive’, albaricoque ‘apricot’, alcachofa ‘artichoke’, alfalfa ‘alfalfa’, algodón ‘cotton’, arroz ‘rice’, azúcar ‘sugar’, limón ‘lemon’, naranja ‘orange’, jazmín ‘jasmine’, alcohol ‘alcohol’, cero ‘zero’, cifra ‘cipher’, cenit ‘zenith’, albóndiga ‘meat ball’, azul ‘blue’, matar ‘to kill’ (Arabic mat ‘dead, checkmate’), mono ‘monkey’, ojalá ‘if Allah will (oh I wish)’, res ‘cattle’. From Taino (Arawakan): canoa ‘canoe’, iguana ‘iguana’, nigua ‘nit’, maíz ‘maize, corn’, ají ‘chili pepper’, yuca ‘sweet manioc’, tuna ‘fruit of prickly pear cactus’, barbacoa ‘barbecue’, batata ‘sweet potato’, enagua ‘petticoat, skirt, native skirt’, huracán ‘hurricane’, sabana ‘savanna’, macana ‘club’, cacique ‘chief’; bejuco ‘vine’, maní ‘peanut’. From Carib: caníbal ‘cannibal’, manatí ‘manatee (sea cow)’, loro ‘parrot’, colibrí ‘hummingbird’, caimán ‘cayman, alligator species’, caribe ‘Carib’, ‘Caribbean’. From Nahuatl: hule ‘rubber’, tiza ‘chalk’, petaca ‘covered hamper, trunk, suitcase’, coyote ‘coyote’, ocelote ‘ocelot’, sinsonte ‘mocking bird’, guajolote ‘turkey’, chocolate ‘chocolate’, cacao ‘cacao, cocoa’, chicle ‘gum, chicle’, tomate ‘tomato’, aguacate ‘avocado’, cacahuete ‘peanut’, tamal ‘tamale’, jícara ‘gourd cup, small gourd bowl’, metate ‘quern, grinding-stone’, mecate ‘string, twine’, pulque ‘pulque (drink from century plant juice)’, achiote ‘bixa (food dye)’, camote ‘sweet potato’, ayote ‘pumpkin’, chayote ‘chayote (a vegetable)’, elote ‘ear of corn’, nopal ‘prickly pear cactus’, guacamole ‘guacamole’, cuate ‘buddy, twin’, caite ‘sandal’. From Quechua: pampa ‘pampa’, papa ‘potato’, coca ‘coca’, quino ‘quinine’, mate ‘mate (a strong tea)’, guano ‘guano (bird fertilizer)’, llama ‘llama’, vicuña ‘vicuña’ (llama species), alpaca ‘alpaca’ (llama species), cóndor ‘condor’, inca ‘Inca’, gaucho ‘gaucho’ (cowboy/ horseman). From Tupi-Guarani: jaguar ‘jaguar’, piraña ‘piranha’ (violent fish), tapioca ‘tapioca’, ananás ‘pineapple’. From English: bistec ‘beefsteak’, ron ‘rum’, huisqui/whisky ‘whisky’, orange crush ‘Orange Crush (a soft drink)’, sandwich/sanduche/sanguich ‘sand-wich’, panqueque ‘pancake’, lonche ‘lunch’, boicot/boicotear ‘boycott’, clip ‘paperclip’, piqueteo ‘picketing’/piquetear ‘to picket’, yate ‘yacht’, parquear ‘to park’, parqueo ‘parking place’, bumper/bómper ‘car bumper’, jet ‘jet’, stop ‘stop’, jeep ‘jeep’; clóset ‘water closet, toilet’, plywood/plaiwud ‘plywood’, álbum ‘album’, bar ‘bar’, film(e)/filmar ‘film’/‘to film’, show ‘show’, ticket/tiquete ‘ticket’, sex appeal/sexapil ‘sex appeal’, stress/estrés ‘stress’, spray/espréi ‘spray’, chequear/checar ‘to check’. (For some of these and for further examples, see Campbell 1997a; Corominas and Pascal1980; Lapesa 1981; Resnik 1981; Spaulding 1965.)
What historical and cultural inferences might you suggest about the nature of the contact between speakers of Spanish and each of these other languages based on these? Concentrate on the Germanic and Arabic contacts.
From Germanic: Germanic speakers had influence in the military sphere, based on such borrowed terms as: bandera ‘flag’, botín ‘booty’, dardo ‘dart’, espiar ‘to spy’, guardar ‘guard’, guerra ‘war’, yelmo ‘helmet’, and hacha ‘axe’. There was some influence in clothing: ropa ‘clothing’, falda ‘skirt’, guante ‘glove’, and toalla ‘towel’. Some colour terms were borrowed: blanco ‘white’ and gris ‘grey’. There was influence in terms for rank, wealth, and commerce: ganar ‘to gain, win, earn’, barón ‘baron’, and rico ‘rich’. The terms for directions were borrowed: norte
39 ‘north’, sur ‘south’, este ‘east’, oeste ‘west’ (note that some of these co-exist along side non- Germanic words for directions: occidente (also poniente) ‘west’, and oriente ‘east’).
From Arabic: (NOTE: many of these loans have incorporated in them the Arabic article al- ‘the’.) Arabic speakers influenced Spain in the following domains: Government, cities, towns, and building: alcázar ‘castle’ (modification of Latin castrum ‘building or plot of land for military defense’, with the Arabic article al-), alcalde ‘mayor’, aldea ‘village’, almacén ‘storehouse’, barrio ‘district of city’, adobe (sun-dried brick), albañil ‘mason’, alcoba ‘bedroom’ (alcove), Household items: almohada ‘pillow’, alfiler ‘pin’, alfombra ‘rug’, Military: alférez ‘ensign’, atalaya ‘watchtower’, bazar ‘bazaar’, Foods and plants: aceite ‘oil’, aceituna ‘olive’, albaricoque ‘apricot’, alcachofa ‘artichoke’, arroz ‘rice’, azúcar ‘sugar’, limón ‘lemon’, naranja ‘orange’, albóndiga ‘meatball’, alfalfa ‘alfalfa’, algodón ‘cotton’, jazmín ‘jasmine’, Mathematics and learning: cero ‘zero’, cifra ‘cipher’, and cenit ‘zenith’.
From Taino: The indigenous languages that Columbus encountered contributed to Spanish terms mostly for new foods and food plants not previously known: maíz ‘maize, corn’, ají ‘chili pepper’, yuca ‘sweet manioc’, tuna ‘fruit of prickly pear cactus’, barbacoa ‘barbecue’, batata ‘sweet potato’, and maní ‘peanut’. A few items that reflect the culture and items important to speakeres of Taino were aslo borrowed: cacique ‘chief’, canoa ‘canoe’, macana ‘club’, enagua ‘petticoat, skirt, native skirt’, sabana ‘savanna’, etc. A few other plants, animals, and other things first encountered in the New World: iguana ‘iguana’, nigua ‘nit’, huracán ‘hurricane’, bejuco ‘vine’.
From Carib: Spanish borrowed terms for newly encountered animals: manatí ‘manatee (sea cow)’, loro ‘parrot’, colibrí ‘hummingbird’, caimán ‘cayman (alligator species’). For people Columbus encountered: caníbal ‘cannibal’, caribe ‘Carib’, ‘Caribbean’. Note that caníbal and caribe have the same origin. Columbus wrote that in the islands he discovered there was great “Caribs” who were called “Caniba” in some islands but were called “Carib” on Hispaniola. They were said to eat anyone they captured.
From Nahuatl: Spanish was heavily influence by Nahuatl in several spheres, including: Foods and food plants: guajolote ‘turkey’, chocolate ‘chocolate’, cacao ‘cacao, cocoa’, chicle ‘gum, chicle’, tomate ‘tomato’, aguacate ‘avocado’, cacahuete or cacahuate ‘peanut’, tamal ‘tamale’, pulque ‘pulque (drink from century plant juice)’, achiote ‘bixa (food dye)’, camote ‘sweet potato’, ayote ‘pumpkin’, chayote ‘chayote (a vegetable)’, elote ‘ear of corn’, nopal ‘prickly pear cactus’, guacamole ‘guacamole’. New World animals: coyote ‘coyote’, ocelote ‘ocelot’, sinsonte ‘mocking bird’. Material culture: petaca ‘covered hamper, trunk, suitcase’, jícara ‘gourd cup, small gourd bowl’, metate ‘quern, grinding-stone’, mecate ‘string, twine’, hule ‘rubber’, tiza ‘chalk’,
40 caite ‘sandal’.
From Quechua: Quechua influence on Spanish was primarily in the area of new plants and animals. New World plants: papa ‘potato’, coca ‘coca’, quino ‘quinine’, mate ‘mate (a strong tea)’, New World animals: llama ‘llama’, vicuña ‘vicuña’ (llama species), alpaca ‘alpaca’ (llama species), cóndor ‘condor’, guano ‘guano (bird fertilizer)’
From Tupi-Guarani: New World animals, plants, and food: jaguar ‘jaguar’, piraña ‘piranha’ (fish species), ananás ‘pineapple’, tapioca ‘tapioca’.
From English: English has had and continues to have influence on Spanish in a number of areas; those represented in this list include terms pertaining to: Foods and drinks: bistec ‘beefsteak’, sandwich/sanduche/sanguich ‘sandwich’, panqueque ‘pancake’, lonche ‘lunch’, ron ‘rum’, huisqui/whisky ‘whisky’, orange crush ‘Orange Crush (a soft drink)’. Automobiles and airplanes: parquear ‘to park’, parqueo ‘parking place’, bumper/bómper ‘car bumper’, jet ‘jet’, stop ‘stop’, jeep ‘jeep’. Arts and entertainment: film(e)/filmar ‘film’/‘to film’, show ‘show’, ticket/tiquete ‘ticket’, sex appeal/sexapil ‘sex appeal’, álbum ‘album’. Economics and labor: boicot/boicotear ‘boycott’, piqueteo ‘picketing’/piquetear ‘to picket’
Which of the non-Germanic words do you think were further borrowed later from Spanish to English (or from Spanish to French and then on to English)? From Arabic: barrio ‘district of city’, adobe (sun-dried brick), alcoba ‘bedroom’ > alcove, bazar ‘bazaar’, albaricoque ‘apricot’ > apricot, alcachofa ‘artichoke’ > artichoke (via Italian), alfalfa ‘alfalfa’, algodón ‘cotton’ > cotton, azúcar ‘sugar’, limón ‘lemon’ > lemon, naranja ‘orange’ > orange (via French), jazmín ‘jasmine’ > jasmine, alcohol ‘alcohol’, cero ‘zero’ > zero, cifra ‘cipher’ > cipher, cenit ‘zenith’ > zenith, albóndiga ‘meat ball’, azul ‘blue’, From Taino (Arawakan): canoa ‘canoe’ > canoe, iguana ‘iguana’, maíz ‘maize, corn’ > maize, barbacoa ‘barbecue’ > barbacue (via French), batata ‘sweet potato’ > potato, From Carib: caníbal ‘cannibal’ > cannibal, manatí ‘manatee (sea cow) > manatee’, caimán ‘cayman > cayman, caribe ‘Carib’, ‘Caribbean’ > Carib, Caribbean. From Nahuatl: coyote ‘coyote’, ocelote ‘ocelot’, chocolate ‘chocolate’, cacao ‘cacao, cocoa’ > cacao, cocoa, chicle ‘gum, chicle’ > chile, tomate ‘tomato’ > tomato, aguacate ‘avocado’ > avacodo, tamal ‘tamale’ > tamale, metate ‘quern, grinding-stone’, pulque ‘pulque (drink from century plant juice)’, guacamole ‘guacamole’. From Quechua: pampa ‘pampa’, coca ‘coca’ > coca, cocaine, quino ‘quinine’ > quinine, mate ‘mate (a strong tea)’, guano ‘guano (bird fertilizer)’, llama ‘llama’, vicuña ‘vicuña’ (llama species), alpaca ‘alpaca’ (llama species), cóndor ‘condor’ > condor, inca ‘Inca’, gaucho ‘gaucho’ (cowboy/horseman). From Tupi-Guarani: jaguar ‘jaguar’, piraña ‘piranha’ > piranha, tapioca ‘tapioca’.
41
Exercise 3.5 Loanwords in Japanese
The following is a list of some of the loanwords into Japanese, primarily from English (though some other European languages may be involved in a few of these). How has Japanese modified the foreign sounds to fit its phonology? What arguments can you make to show that the direction of borrowing is indeed from English into Japanese? State your evidence. NOTE: Japanese permits no syllable-final consonants other than -n; it does not tolerate consonant clusters other than -nC, though geminates [double consonants] are allowed, and the only word-final consonant is -n. In Japanese, /t/ is [ts] before u, [č] ([tʃ]) before i, and [t] elsewhere; similarly, /s/ is [š] ([ʃ]) before i. Japanese has no l or v, and no h before u (only f), and no ə.
1. aidea ‘idea’ 2. aidoru ‘idol’ (celebrity) 3. airon ‘iron’ (appliance) 4. aisukurīmu ‘ice cream’ 5. amachua ‘amateur’ 6. ampaia ‘umpire’ 7. asuterisuku ‘asterisk’ 8. ba ‘bar’ 9. baffaro ‘buffalo’ 10. bagen ‘bargain, sale’ 11. baiburu ‘Bible’ 12. baiorin ‘violin’ 13. baitaritī ‘vitality’ 14. bajji ‘badge’ 15. baketsu ‘bucket’ 16. ban ‘van’ 17. bando ‘band, belt’ 18. baraddo ‘ballad’ 19. barē-boru ‘volleyball’ 20. basuketto ‘basket’ 21. basu-taminaru ‘bus terminal’ 22. bata ‘butter’ 23. batterī ‘battery’ 24. batto ‘(baseball) bat’ 25. bēju ‘beige’ 26. bekon ‘bacon’ 27. benchi ‘bench’ 28. beruto ‘belt’ 29. bifuteki ‘beefsteak’ 30. bīru ‘beer’ 31. bīrusu ‘virus’ (cf. uirusu) 32. bitamin ‘vitamin’
42 33. bōi sukauto ‘Boy Scout’ 34. borantia ‘volunteer’ 35. boru ‘ball’ 36. boru ‘bowl’ (stadium) 37. borutto ‘bolt’ (headed metal pin) 38. bosu ‘boss’ 39. botan ‘button’ 40. bumu ‘boom’ (in prosperity) 41. burajā ‘bra’ (< brassiere) 42. burausu ‘blouse’ 43. buresuretto ‘bracelet’ 44. burijji ‘bridge’ (card game) 45. burondo ‘blond’ 46. buronzu ‘bronze’ 47. chata ‘charter’ 48. chato ‘chart’ 49. channeru ‘channel’ 50. chansu ‘chance’ 51. chīfu ‘chief’ 52. chīmuwāku ‘teamwork’ 53. chīzu ‘cheese’ 54. chokorēto ‘chocolate’ 55. chūbu ‘tube’ 56. daietto ‘diet’ 57. daiyamondo ‘diamond’ 58. dansu ‘dance’ 59. depato ‘department store’ 60. dezato ‘dessert’ 61. doa ‘door’ 62. doraggu sutoa ‘drugstore’ 63. doresu ‘dress’ 64. doru ‘dollar’ 65. epuron ‘apron’ 66. erebeta ‘elevator, lift’ 67. fan ‘fan’ (admirer) 68. firumu ‘film’ 69. fōku ‘fork’ 70. fūdo ‘hood’ 71. furaipan ‘frying pan’ 72. furokku-koto ‘frock-coat’ 73. furutsu jusu ‘fruit juice’ 74. gādoru ‘girdle’ 75. garasu ‘glass, pane’ 76. garēji ‘garage’ 77. garon ‘gallon’ 78. gāru sukauto ‘Girl Scout’
43 79. gasorin ‘gasoline’ 80. gāta ‘garter’ 81. gēmu ‘game’ 82. gōru ‘goal’ 83. gorufu ‘golf’ 84. goshippu ‘gossip’ 85. gureibī ‘gravy’ 86. gurōbu ‘glove’ 87. gyamburu ‘gamble’ 88. gyaroppu ‘gallop’ 89. hābu ‘herb’ 90. hado-wea ‘hardware’ (computer) 91. hambāga ‘hamburger’ 92. hamma ‘hammer’ 93. hamu ‘ham’ 94. hamueggu ‘ham and eggs’ 95. handoru ‘handle, steering wheel’ 96. hankachi ‘handkerchief’ 97. herumetto ‘helmet’ 98. hinto ‘hint’ 99. hitto ‘hit’ 100. hotto doggu ‘hotdog’ 101. hyuzu ‘fuse’ 102. inchi ‘inch’ 103. indekkusu ‘index’ 104. infuruenza ‘flu’ 105. īsuto ‘yeast’ 106. jāji ‘jersey’ 107. jakketto ‘jacket’ 108. jamu ‘jam’ 109. jampa ‘jumper’ 110. jazu ‘jazz’ 111. jigu-zagu ‘zigzag’ 112. jīnzu ‘jeans’ 113. jippa ‘zipper’ 114. kādo ‘card’ 115. karu ‘curl’ 116. katen ‘curtain’ 117. katon ‘carton’ 118. kan ‘can, tin’ 119. kareji ‘college’ 120. katarogu ‘catalogue’ 121. kaunta ‘counter’ 122. kēki ‘cake’ 123. kisu ‘caress, kiss’ 124. kochi ‘coach’ (trainer)
44 125. kōhī ‘café, coffee’ 126. kokku ‘cook’ 127. komasharu ‘commercial’ 128. komedī ‘comedy’ 129. kompakuto disuku ‘compact disk’ 130. kompyuta ‘computer’ 131. komyunikeshon ‘communication’ 132. konkurīto ‘concrete’ (building material) 133. kopī ‘copy’ 134. koppu ‘cup’ 135. kurabu ‘club’ 136. kuraimakkusu ‘climax’ 137. kuriketto ‘cricket’ (game) 138. kyabetsu ‘cabbage’ 139. kyabia ‘caviar’ 140. kyampasu ‘campus’ 141. kyampen ‘campaign’ 142. kyampu ‘camp’ 143. kyandī ‘candy’ 144. kyaputen ‘captain’ (chief, leader) 145. māchi ‘march’ 146. māketto ‘market’ 147. māmarēdo ‘marmalade’ 148. manēja ‘manager’ 149. maton ‘mutton’ 150. memba ‘member’ 151. merodī ‘melody’ 152. meron ‘melon’ 153. mineraru uota ‘mineral water’ 154. miruku ‘milk’ 155. mishin ‘sewing machine’ 156. morutaru ‘mortar’ 157. myūjikaru ‘musical’ (play) 158. naifu ‘knife’ 159. namba ‘number’ 160. napukin ‘napkin’ 161. nattsu ‘nut’ 162. nekutai ‘necktie’ 163. nettowāku ‘network’ 164. nikkeru ‘nickel’ 165. nūdoru ‘noodle’ 166. nyūsu ‘news’ 167. ōba ‘overcoat’ 168. ōbun ‘oven’ 169. ōke ‘O.K.’ 170. ōkesutora ‘orchestra’
45 171. ofisu ‘office’ 172. omuretsu ‘omelette’ 173. orenji ‘orange’ 174. oribu ‘olive(s)’ 175. pai ‘pie’ 176. painappuru ‘pineapple’ 177. painto ‘pint’ 178. paionia ‘pioneer’ 179. paipu ‘pipe’ 180. paipu-orugan ‘organ’ (musical instrument) 181. panchi ‘punch’ (fruit punch) (cf. also ponchi, ponsu) 182. paneru ‘panel’ 183. panfuretto ‘pamphlet’ 184. panikku ‘panic’ 185. pantī ‘panties’ 186. pantsu ‘underpants, shorts, drawers’ 187. parēdo ‘parade’ 188. parupu ‘pulp (wood-pulp)’ 189. pāsento ‘percent’ 190. pasupōto ‘passport’ 191. pātī ‘party’ 192. patoron ‘patron’ 193. paturōru ‘patrol’ 194. pējento ‘pageant’ 195. pedaru ‘pedal’ 196. pen ‘pen’ 197. pēsuto ‘paste’ 198. perikan ‘pelican’ 199. pikunikku ‘picnic’ 200. pin ‘pin’ 201. pīnatsu ‘peanut’ 202. pinku ‘pink’ 203. piru ‘pill’ 204. pisutoru ‘pistol’ 205. poketto ‘pocket’ 206. pondo ‘pound’ 207. ponsu ‘punch‘ (fruit punch) (cf. also panchi, ponchi) 208. posutā ‘poster’ 209. posuto ‘mailbox, postbox’ 210. puragu ‘plug’ (electric) 211. puramu ‘plum’ 212. purasuchikku ‘plastic’ 213. purin ‘pudding’ (cf. also pudingu) 214. pūru ‘pool’ 215. rādo ‘lard’ 216. raberu ‘label’
46 217. raifuru- ‘rifle’ 218. raimu ‘lime’ 219. rain ‘line’ 220. rajieta ‘radiator’ 221. rajio ‘radio’ 222. ramu ‘rum’ 223. rampu ‘lamp’ 224. rantan ‘lantern’ 225. raunji ‘lounge’ 226. rejisuta ‘register’ 227. rekōdo ‘record’ 228. repōto ‘report’ 229. renji ‘stove’ (cooking stove, < ‘range’) 230. rēsu ‘lace’ 231. renzu ‘lens’ 232. resuringu ‘wrestling’ 233. retasu ‘lettuce’ 234. ribon ‘ribbon’ 235. rinen ‘linen’ 236. risuto ‘list’ 237. rizoto ‘resort’ 238. rizumu ‘rhythm’ 239. romansu ‘romance’ 240. rosuto chikin ‘roast chicken’ 241. sākasu ‘circus’ 242. sain ‘sign, signal, signature’ 243. sararī ‘salary’ 244. sāroin ‘sirloin’ 245. sekkusu ‘sex’ 246. serori ‘celery’ 247. sēta ‘sweater’ 248. shaberu ‘shovel’ 249. shatsu ‘shirt’ (cf. waishatsu ‘dress shirt’ < ‘white shirt’) 250. sheruta ‘shelter’ 251. shiringu ‘shilling’ 252. shiroppu ‘syrup’ 253. shisuta ‘sister’ 254. shīzun ‘season’ 255. sho ‘show’ (entertainment) 256. shokku ‘shock’ 257. shoru ‘shawl’ 258. sokkusu ‘socks’ 259. sōda ‘soda’ 260. sōsēji ‘sausage’ 261. suchimu ‘steam’ 262. suchuwādo ‘steward’ (aeroplane)
47 263. suchuwadesu ‘stewardess’ 264. sukejuru ‘schedule’ 265. sukurīn ‘screen’ (movie screen) 266. sukuryu ‘screw’ 267. sukyandaru ‘scandal’ 268. sukato ‘skirt’ 269. sumāto ‘smart’ (fashionable) 270. supaisu ‘spice’ 271. superu ‘spelling’ 272. supīka ‘speaker’ (loudspeaker) 273. suponji ‘sponge’ 274. supōtsu ‘sport’ 275. sūpu ‘soup’ 276. supūn ‘spoon’ 277. supurē ‘spray’ 278. supurinkura ‘sprinkler’ 279. surakkusu ‘slacks’ 280. surangu ‘slang’ 281. surippa ‘slipper(s)’ 282. surōgan ‘slogan’ 283. sutā ‘star’ (film star) 284. sutairu ‘style’ 285. sutajio ‘studio’ 286. sutēki ‘steak’ 287. sutenresu ‘stainless steel’ 288. sutōbu ‘stove’ (heating stove) 289. sutoraiki ‘strike’ (by employees) 290. sutoraiku ‘strike’ (in baseball) 291. sūtsukesu ‘suitcase’ 292. sutsūru ‘stool’ 293. taipuraita ‘typewriter’ 294. tairu ‘tile’ 295. taiya ‘tyre’ 296. takkuru ‘tackle’ (in football) 297. takushi ‘cab, taxi’ 298. tambura ‘tumbler’ (drinking glass) 299. tātorunekku ‘turtleneck’ 300. tēburu ‘table’ 301. tenisu ‘tennis’ 302. terebi(jon) ‘television’ 303. tīn-eija ‘teenager’ 304. toire ‘toilet’ (lavatory) 305. ton ‘ton’ 306. torakku ‘truck’ 307. torikku ‘trick’ 308. tōsuto ‘toast’
48 309. tsuīdo ‘tweed’ 310. ueita ‘waiter’ 311. ueitoresu ‘waitress’ 312. uesuto ‘waist’ 313. uīkuendo ‘weekend’ 314. uinku ‘wink’ 315. uirusu ‘virus’ (cf. bīrusu) 316. wain ‘wine’ 317. wakuchin ‘vaccine’ 318. wanisu ‘varnish’ 319. yādo ‘yard (measure)’ 320. yotto ‘yacht’ 321. yunifomu ‘uniform’
How has Japanese modified the foreign sounds to fit its phonology? English vowels /a/, /æ/, /ʌ/ or /ǝ/, and /ɚ/ (the last spelled
49 Japanese breaks up the consonant clusters in the English loans by inserting a vowel, usually /u/, as in: 297. takushi ‘taxi’, and in 265. sukurīn ‘screen’ (which has two inserted vowels to break up the three consonants of the cluster, skr > sukur... (see also 7., 28.-29., 37., 42.-46., 73., 261.-292., etc.). [NOTE: this u is often voiceless in many of these words, depending on the phonological environment in which it appears.] There is a tendency for plain voiceless stops in English to be replaced by geminate voiceless stops between short vowels, as in: 20. basuketto ‘basket’, 84. goshippu ‘gossip’, 103. indekkusu ‘index’ (see also: 24., 37., 43., 62., 72., 88., 97., 100., 307., etc.). This is especially true if the English word is spelled with double consonants, as in 281. surippa ‘slippers’, or 306. torakku ‘truck’ (with ck). Following Japanese phonological patterns, alveolar stops and fricatives are palatalized to affricates before /i/, as in: 52. chīmuwāku ‘teamwork’, 111. jigu-zagu ‘zigzag’, 253. shisuta ‘sister’, and 254. shīzun ‘season’ (also in 84., 106., 113., 157., 212., 220., 252., 261., 264., 285., 302., 317., etc.). In Japanese, /t/ is [ts] before /u/, seen in 292. sutsūru ‘stool’ (also in 15., 73., 138., 161., 172., 201., 249., 274., 291., 309.). Note that the /č/ of 55. chūbu ‘tube’, 262. suchuwādo ‘steward’, and 263. suchuwadesu ‘stewardess’ reflect not the /tu/ English of most varieties of American English, but rather the /tyu/, as in British English in these words. Also Japanese has /s/ before /u/, leading English words with /š/ (IPA [ʃ]) to be borrowed with /s/ in this environment in the loanwords, as in 318. wanisu ‘varnish’, see also 207. Note that some of the English loans do not conform to the distribution of the native sounds in Japanese.
What arguments can you make to show that the direction of borrowing is indeed from English into Japanese? State your evidence. Morphological complexity clues – a good number of these forms are polymorphemic in English, but monomorphemic in Japanese, suggesting the English forms, with their internal morphological analysis, are the sources of these loans, for example: 4. aisukurīmu ‘ice cream’, 13. baitaritī ‘vitality’, 52. chīmuwāku ‘teamwork’, 66. erebeta ‘elevator’, 71. furaipan ‘frying pan’, 94. hamueggu ‘ham and eggs’, 96. hankachi ‘handkerchief’, 112. jīnzu ‘jeans’, 130. kompyuta ‘computer’, 131. komyunikeshon ‘communication’, etc. Geographical and ecological clues – things not natively found in Japan, foreign to Japan: 9. baffaro ‘buffalo, 54. chokorēto ‘chocolate, 125. kōhī ‘coffee’, 174. oribu ‘olive’, 176. painappuru ‘pineapple’, 201. pīnatsu ‘peanut’, 251. shiringu ‘shilling. Phonological clues – forms that contain sounds not native to Japanese or which violate Japanese phonological patterns: 5. amachua ‘amateur’ (with ch instead of expected ts before u); 25. bēju ‘beige’ (with j before u); 34. borantia ‘volunteer’, (with t before i instead of expected ch); 47. chata ‘charter’, 48. chato ‘chart’, 49. channeru ‘channel’, 50. chansu ‘chance, 55., 262., 263. (with ch before a vowel other than i); 127. komasharu ‘commercial’, 131. komyunikeshon ‘communication’, 248.-250., 255.-257. (with sh before a vowel other than i). Other semantic clues – the clues to the direction of change for some of the loans involve various semantic considerations, some perhaps associated with one of the other categories already mentioned (geographical and ecological clues). For example, we might assume 11. baiburu ‘Bible’ is a loan because Christianity and the Bible are not associated with traditional Japanese culture and history. Names for or terms involving various sports and games known to
50 have their origins in Europe or America might be suspected to be borrowings: 19. barē-boru ‘volleyball’, 24. batto ‘bat (baseball)’, 36. boru ‘bowl’ (stadium), 44. burijji ‘bridge’ (card game), 83. gorufu ‘golf, 137. kuriketto ‘cricket’ (the game), 296. takkuru ‘tackle’ (in football), 290. sutoraiku ‘strike’ (in baseball ), 301. tenisu ‘tennis. Others with known or suspected origins elsewhere might include: 29. bifuteki ‘beefsteak’, 30. bīru ‘beer’, 33. bōi sukauto ‘Boy Scout’, 91. hambāga ‘hamburger, 78. gāru sukauto ‘Girl Scout’, 100. hotto doggu ‘hotdog’, 112. jīnzu ‘jeans’, etc.
Exercise 3.6 Hispanisms in Mayan languages
The following is a list of some of the ‘hispanisms’ (loanwords from Spanish) found in some of the Mayan languages (of Mexico and Guatemala). The Spanish forms are presented both in current pronunciation and in that of the sixteenth century. Based on these, what evidence can you derive from these loans in the Mayan languages relevant to changes which have taken place in Spanish since these forms were borrowed? By way of illustration, consider the following example involving Sayula Popoluca (a Mixe-Zoquean language): Spanish caja ‘box’ (modern [kaxa], colonial [kaSa]: Sayula Popoluca kaʃa ‘coffin’ (‘box for the dead’). From this, you would tentatively conclude that Spanish has undergone the change of ʃ > x after this word was borrowed. Of course, it is necessary to keep in mind that the borrowing language will make substitutions, replacing the Spanish sounds with the closest phonetic counterpart available in the recipient language, so that not all differences in the borrowing language will be due to changes which Spanish has subsequently undergone; to determine this, it will be necessary to compare the sixteenth-century and the modern Spanish forms. In regard to this particular example, it is interesting that Sayula Popoluca later borrowed the Spanish word for ‘box’ again, after the change, as kaha ‘cardboard box’ (note Sayula Popoluca has no [x] so that [h] is the language’s closest approximation to Modern Spanish [x]). Note the following phonetic symbols found in these examples: [s]9 dental (fronted) s [ṣ] apical alveolar s [ṣ̌] laminal retroflex š
j Focus on /l / and /j/: j j 1. llave ‘key’ (modern [jaβe], colonial [l aβe, l ave]): Akateko laweh, Q’anjobal lawe, K’iche’ lawe. j 2. cebolla ‘onion’ (modern [seβoja], colonial [s9ebol a]): Akateko sewolya, Q’anjobal j sewolia, Tzeltal sebolia (none of the Mayan languages has /l /, but they do have /l/ and /j/). j 3. cuchillo ‘knife’ (modern [kučijo], colonial [kučil o]): Chol kučilu, Huastec kuči:l, Q’anjobal kučiilu ‘knife, razor’, K’iche’ kučiʔl. j 4. silla ‘chair’ (modern [sija], colonial ṣil a]): Akateko ʃilah, Huastec ʃi:laʔ, ‘saddle, chair’, Q’anjobal ʃila, K’iche’ ʃila, Tzotzil ʃila.
51 j 5. castellano ‘Castilian, Spanish (modern [kastejano], colonial [kaṣtel ano])’: Choltí kaʃtilan čaɓ ‘sugar’ (literally ‘Castilian honey’), kaʃtilan wa ‘bread’ (literally ‘Castilian tortilla’); K’iche’ kaʃtilan, kaʃlan ‘Castilian, Spanish, pure, correct’.
Focus on /s̯/, /ṣ/, and /ʃ/: 6. sartén ‘frying pan’ (modern [saRtén], colonial [ṣaRtén]): Q’anjobal ʃalten, ʃaltin, Motocintlec ʃarten, Tzotzil ʃalten. 7. sebo ‘tallow, fat’ (modern [seβo], colonial [ṣebo]): Q’anjobal ʃepuʔ, K’iche’ ʃepu, ʃepo, Tzotzil ʃepu. 8. seda ‘silk’ (modern [seða], colonial [ṣeða]): Chol ʃelah ‘ribbon’, Tzotzil ʃela ‘silk, ribbon’. (Mayan languages have no [ð].) 9. semana ‘week’ (modern [semana], colonial [ṣemana]): Q’eqchi’ ʃama:n, ʃema:n, K’iche’ ʃemano, Tzotzil ʃemana. j j 10. señora ‘lady, madam, mrs.’ (modern [sen ora], colonial [ṣen ora]): Chol ʃinolah ‘non- Indian woman’, Mam ʃnu:l ‘non-Indian woman’, Motocintlec ʃnu:la:n ‘non-Indian woman’, Tzeltal ʃinola ‘non-Indian woman’. 11. mesa ‘table’ (modern [mesa], colonial [meṣa]): Akateko meʃah, Huastec me:ʃa, Q’eqchi’ me:ʃ, Motocintleco me:ʃah, K’iche’ meʃa. 12. patos ‘ducks’ (modern [patos], colonial [patoṣ]): Huastec pa:tuʃ, Q’eqchi’ patuʃ, K’iche’ pataʃ, Tzotzil patoʃ. (Note that several plant and animal terms, though singular, were borrowed from the Spanish plural form, as in this example and the next.) 13. vacas ‘cows’ (modern [bakas], colonial [βakaṣ, vakaṣ]: Akateko wakaṣ̌ ‘cattle’, Chol wakaʃ ‘bull, cow’, Itzá wakaʃ ‘cattle’, Q’anjobal wakaṣ̌ ‘cow, cattle’, Q’eqchi’ kwakaʃ ‘cow, cattle’, Mopan wakaʃ ‘cow, bull, cattle’, Tzeltal wakaʃ ‘beef’. (See also 4. and 5. above.) 14. cidra ‘ a grapefruit-like fruit’ (modern [siðra], colonial [si̯ ðra]): Chol silah, Tzotzil sila. (Note that these languages have no d, ð, or r). 15. cocina ‘kitchen’ (modern [kosina], colonial [kosina̯ ]): Motocintleco kusi:nah, Tzotzil kusina. 16. cruz ‘cross’ (modern [krus], colonial [krus]:̯ Chol rus, Q’anjobal kurus, Q’eqchi’ kurus, Mam lu:s, Motocintleco kuru:s, Tzotzil kurus. 17. lazo ‘lasso, rope’ (modern [laso], colonial [laso̯ ]): Akateko lasuh, Chol lasoh, Tzeltal laso, Tzotzil lasu. 18. taza ‘cup’ (modern [tasa], colonial [tasa̯ ]): Chol tasa ‘piece of glass’, Huastec ta:sa, Q’eqchi’ ta:s. 19. jabón ‘soap’ (modern [xaβón], colonial [ʃabón]): Chol ʃapum, ʃapom, Huastec ʃabu:n, *Jakalteko ṣ̌apun, *Q’anjobal ṣ̌apun, Motocintlec ʃa:puh, K’iche’ ʃɓon, Tzeltal ʃapon. 20. jarro ‘jug, jar’ (modern [xaro], colonial [ʃaro]): Jakalteko ʃalu, Q’anjobal ʃalu, Mam ṣ̌ar, Motocintleco ʃa:ruh, K’iche’ ʃaruʔ, Tzeltal ʃalu, Tzotzil ʃalu. 21. aguja ‘needle’ (modern [aguxa], colonial [aguʃa]): Akateko akuʃah, Chol akuʃan, Q’anjobal akuʃa, Q’eqchi’ aku:ʃ, ku:ʃ, Motocintlec aku:ʃah, Tzeltal akuʃa, Tzotzil akuʃa. 22. caja ‘box’ (modern [kaxa], colonial [kaʃa]): Chol kaʃa-teʔ ‘chest’ (teʔ = ‘wood’), Q’anjobal kaʃa ‘box, chest’, Q’eqchi’ ka:ʃ ‘chest’, Mam ka:ṣ̌ ‘box’, Motocintlec ka:ʃah ‘box, chest’, K’iche’ kaʃa ‘box, chest, trunk’, Tzeltal kaʃa.
52 Focus on /v/ and /b/: 23. ventana ‘window’ (modern [bentana], colonial [βentana, ventana]): Chol wentana, Q’anjobal wentena, Motocintlec wanta:nah. 24. (= 13. above) vacas ‘cows’ (modern [bakas], colonial [βakaṣ, vakaṣ]: Akateko wakaṣ̌ ‘cattle’, Chol wakaʃ ‘bull, cow’, Itzá wakaʃ ‘cattle’, Q’anjobal wakaṣ̌ ‘cow, cattle’, Q’eqchi’ kwakaʃ ‘cow, cattle’, Mopan wakaʃ ‘cow, bull, cattle’, Tzeltal wakaʃ ‘beef’. 25. calvario ‘Calvary’ (modern [kalβaRio], colonial [kalβaRio, kalvaRio]) : Q’anjobal karwal ‘cemetery, graveyard’, K’iche’ kalwar. 26. clavos ‘nails’ (modern [klaβos], colonial [klaβoṣ, klavoṣ]: Akateko lawuʃ, Chol lawuʃ, Tzeltal lawuʃ, Tojolabal lawuʃ. (Note, these forms mean ‘nail’, but are borrowed from the Spanish plural form.) 27. rábanos ‘radishes’ (modern [ráβanos], colonial [ráβanoṣ, rávanoṣ]: Tojolabal lawuniʃ, Motocintleco luwaʔnʃa ‘rábano’, Tzotzil alavanuʃ. (Note, these all mean ‘radish’, though borrowed from the Spanish plural form. Tzotzil has a phonemic contrast between /v/ and /b/, but has no /w/; the other languages have no /v/, but do have /w/.) (See also 1.) 28. botón(es) ‘button(s)’ (modern [botón], colonial [botón]: Q’eqchi’ ɓoto:nʃ, K’iche’ ɓotona, ɓotoniʃ, Tojolabal ɓoton ‘button, knot in wood’, Tzotzil ɓoton. 29. bolsa ‘bag, pocket’ (modern [bolsa], colonial [bolṣa, borṣa]): Chol borʃa, Q’eqchi’ ɓo:ʃ ‘pocket’, K’iche’ ɓorʃa, Tzeltal ɓolsa. 30. nabos ‘turnips’ (modern [naβos], colonial [naboṣ]): K’iche’ napuʃ, Tzotzil napuʃ, Motocintlec kolinaʔwa. (See also 2. and 7.)
Based on these, what evidence can you derive from these loans in the Mayan languages relevant to changes which have taken place in Spanish since these forms were borrowed?
Examples 1.-5.: Spanish still had the contrast between lj and j (“y”) when these forms were borrowed. Later these varieties of Latin American Spanish underwent lj > j (merger of lj and j to j).
The examples in 4.-13., 26.-30. with original /ṣ/ and examples in 2., 14.-18. with original /s̯/ show that these two sounds contrasted in the variety of Spanish from which these words were borrowed. Later these varieties of Latin American Spanish underwent ṣ, s̯ > s (merger of ṣ and s̯ with s). Meanwhile, examples 19.-22. do not present a clear picture. They had ʃ in earlier Spanish at the time these were borrowed, but for most of the languages the borrowings do not distinguish this ʃ from ṣ, words with both sounds typically borrowed with ʃ in these languages. However, Q’anjobal, Akateko, and Mam sometimes do have ṣ̌ in loans which originally had ʃ in Spanish, but typically do not have this retroflexed sound in words from Spanish which originally had ṣ. This suggests, but does not conclusively demonstrate, that Spanish at the time these words were borrowed had two distinct sounds. One of these we know from other facts was originally ṣ and shows up in these hispanisms as ʃ; the other was originally ʃ and shows up in these borrowings mostly as ʃ, but in the languages which have a contrast between ʃ and ṣ̌ can show up as ṣ̌, which is not true of words which in Spanish originally had ṣ
Examples 23.-26. show that the variety of Spanish from which these forms were borrowed
53 contrast with examples. 27.-30. The examples in 1., 23.-26. which originally had a fricative, β or v, and examples in 2., 7., 19., 27.-30 with original sotp b show that these two sounds contrasted in the variety of Spanish from which these words were borrowed. The words with original stop b have /b/, /p/, or /ɓ/, that is a stop, in most of these languages, while the words with a fricative (β or v) were mostly borrowed with /w/ in these languages, most of which do not have labial fricatives. Later these varieties of Latin American Spanish underwent the merger of these sounds, β or v and b > /b/ (with [b] and [β] in complementary distribution, [β] between vowels, [b] elsewhere with some variation in contexts involving /VbC depending on the variety and the particular C following it).
54
EXERCISES FOR CHAPTER 4 ANALOGICAL CHANGE
Exercise 4.1 Observe the language of your friends and of newspapers, television and so on, and attempt to find examples of your own of the various sorts of analogy.
No specific answers are offered here – just look for analogies used in the language around you. In slang, advertising, puns, and jokes one usually finds many examples.
Exercise 4.2 Identifying analogical changes
Determine what kind of analogical change is involved in the following examples. Name the kind of change, and attempt to explain how it came about, if you can.
(1) In some dialects of English, the pattern bring/brought/brought has become bring/brang/brung.
Analogical extension: bring/brang/brung is probably modeled on the pattern of verbs such as sing/sang/sung.
(2) Where Standard English has drag/dragged, some varieties of English have drag/drug. It appears in this case that the Standard English pattern is older.
Analogical extension: drag/drug is probably modeled on the pattern of verbs such as dig/dug, hang/hung.
(3) Old Spanish siniestro ‘left’ changed from Latin sinister ‘on the left’ to take ie under the influence of the antonym diestro ‘right’, since diestro and siniestro frequently occurred together.
Immediate model: since sinister ‘left’ and diestro ‘right’ frequently occurred together, the former changed into siniestro by analogy to the latter.
(4) In many Spanish dialects, an intervocalic d is regularly lost, as in mercado > mercao ‘market’; in some instances, however, there are changes of the following sort: dialect bacalado < Standard bacalao ‘codfish’; dialect Bilbado < Standard Bilbao (a place name).
Hypercorrection: since speakers of these Spanish dialects are conscious of the fact that in words were their dialects lack intervocalic d, the more prestigious standard variety has d between the vowels, they insert a d in attempts to sound more like Standard Spanish, but in the process they overgeneralize the insertion of d in some instances and so produce forms which, from the point of view of the standard language, erroneously contain a d in these words where the Standard variety lacks intevocalic d.
55
(5) In the Dominican Republic, forms such as Standard Spanish atras ‘behind’ become astras; in this variety of Spanish, preconsonantal s is often lost, as in ata < asta (spelled hasta) ‘until’.
Hypercorrection: since speakers in the Dominican Republic are conscious of the fact that where their dialect often lacks s before a medial t, Standard Spanish often has s in words such as these, so they insert an s in attempts to sound more like Standard Spanish, but in the process insert s in some words which do not have s in the Standard language, and so produce forms which, from the point of view of the standard language erroneously contain s in words where the Standard variety lacks intevocalic s.
(6) English Jerusalem artichoke (a kind of sunflower, with some similarities to an artichoke) is in origin from Italian girasóle articiocco, where Italian girasole /ǰirasóle/ contains gira- ‘turn around, revolve, rotate’ + sole ‘sun’, and articiocco ‘artichoke’, with nothing associated with Jerusalem; originally.
Folk etymology: because Italian girasole sounds to English speakers something like ‘Jerusalem’, they folk-etymologized it to Jerusalem, even though girasole had nothing to do with Jerusalem.
(7) In English, Key West (in Florida) comes from the Spanish name cayo hueso, where cayo is ‘key, small island’ and hueso is ‘bone’.
Folk etymology: English speakers related Spanish hueso to English west because it sounded somewhat similar and folk-etymologized the name to ‘Key West’.
(8) English heliport < helicopter + airport; snazzy < snappy + jazzy; jumble < jump + tumble.
Blending: two semantically somewhat similar or related words were combined to form the single new words in these cases.
(9) Colloquial and regional varieties of Spanish have haiga where Standard Spanish has haya (subjunctive, ‘there may be’) and vaiga where Standard Spanish has vaya (subjunctive, ‘may go’). These have been influenced by Standard Spanish verb forms such as traiga (subjunctive of traer ‘to bring’, ‘may bring’) and caiga (subjunctive of caer ‘to fall’, ‘may fall’).
Analogical extension: the pattern found in some frequent verbs which in the subjective end in - aiga has been extended to haya and vaya, which formerly were not members of the class of verbs that terminated in -aiga in the subjunctive, changing them to haiga and vaiga, respectively.
(10) Middle English had help- ‘present tense’, holp ‘past tense’; Modern English has help, helped for these.
Analogical leveling: the strong-verb pattern help/holp was leveled on analogy to the regular weak-verb pattern, as in verbs such as bake/baked.
56
(11) English to emote is derived from emotion; to enthuse is derived from enthusiastic.
Back formation: on analogy to words in the pattern of, for example, quote/quotation and promote/promotion, the noun emotion was assumed to be based on a verb root emote plus the suffix -tion, and from this the verb emote was created by backformation. Enthuse is also a backformation, on analogy to the same pattern that gives paraphrase/paraphrastic.
(12) Many varieties of English have a new verb to liaise based on liaison.
Backformation: possibly on analogy to many verbs ending in /-z/ which have related nouns, such as analyze/analysis.
(13) English to diagnose is derived from diagnosis.
Back formation: the new verb diagnose was derived analogically on the model of the many other verb-noun pairs in which a noun ending in -sis has a corresponding verb ending in /-z/, as for example, hypothesize/hypothesis.
(14) Finnish rohtia ‘to dare’ resulted from both rohjeta ‘to be bold enough, to dare’ and tohtia ‘to dare’.
Blending: two semantically related words were combined to form a single new word.
(15) English hangnail is derived from Old English angnaegl ‘painful corn (on foot)’. When ang ‘pain’ as an independent word was lost (though later reborrowed in anguish), the angnaegl form was reinterpreted as having something to do with ‘hanging’, with painful detached skin of toes and then also hands.
Folk etymology: the word came to be associated with hang by folk etymology and thus was reanalyzed as containing hang.
57
EXERCISES FOR CHAPTER 5 THE COMPARATIVE METHOD AND LINGUISTIC RECONSTRUCTION
Exercise 5.1 Aimaran
Consider the following data from the two major branches of the Aimaran language family (Peru and Bolivia). Focus your attention on the sibilant fricatives (s and š) only (ignore x and χ for this exercise). What will you reconstruct? How many sibilant fricatives do you postulate for Proto-Aimaran? State your evidence. NOTE: š = IPA [ʃ], č = IPA [tʃ]; ) = voiceless uvular fricative; C’ = glottalized [ejective] consonants.
Central Aimara Southern Aimara gloss 1. saxu sawu- ‘to weave’ 2. sa(wi) sa(ta) ‘to plant’ 3. asa asa- ‘ to carry flat things’ 4. usu usu- ‘to become sick’ 5. nasa nasa ‘nose’ 6. aski hiskhi ‘to ask’ 7. muxsa muχsa ‘sweet’ 8. suniqi sunaqi ‘small spring’ 9. šanq’a sanqa ‘to snuffle’ 10. waša wasa ‘silent place’ 11. iši isi ‘dress’ 12. muši musi ‘to take care (of)’ 13. puši pusi ‘four’ 14. išt’a hist’a- ‘to close’ 15. išapa isapa- ‘to hear, listen’ (Cerrón-Palomino 2000: 145–6)
Sound correspondences: Central Aimara Southern Aimara Proto-Aimaran s s *s š s *š
Sound change: *š > s in Southern Aimara
What will you reconstruct? *s and *š as distinct phonemes.
How many sibilant fricatives do you postulate for Proto-Aimaran? Two.
58 State your evidence. The two sounds s and š appear both before and after all vowels and there does not appear to be any conditioning from the consonants. There is no plausible phonetic environment which would have conditioned one of these sounds to change to the other in a particular context in these data. The two sounds thus appear to be distinct in Proto-Aimaran.
Exercise 5.2 Tulu
Tulu is a Dravidian language (of India) which has several varieties. Consider the following data from two principal varieties. Focus your attention only on the nasals. What will you reconstruct for these? How many nasals do you postulate for Proto-Tulu? State your evidence. NOTE: j = [ǰ], IPA [dʒ]; ṇ = IPA [ɳ].
Shivalli Sapaliga gloss 1. a:ṇɨ a:nɨ ‘male’ 2. uṇɨ a:nɨ ‘dine’ 3. maṇṇ_ mann_ ‘soil’ 4. ko:ṇε ko:nε ‘room’ 5. e:ṇɨ ya:nɨ ‘I’ 6. ninε ninε ‘wick’ 7. ja:nε da:nε ‘what’ 8. sanε tanε ‘conceiving’ (Bhat 2001: 11)
Sound correspondences: Shivalli Sapaliga Proto-Tulu n n *n ṇ n *ṇ (retroflex) Sound change: *ṇ > n in Sapaliga
What will you reconstruct for these (the nasals)? *n and *ṇ.
How many nasals do you postulate for Proto-Tulu? Two.
State your evidence. Both n and ṇ occur in the same vowel environments and there appears to be no plausible phonetic conditioning for the difference in the two sounds from the other consonants. There is no plausible phonetic environment which would have conditioned one of these sounds to change to the other in these data. The two sounds thus appear to be distinct in Proto-Tulu.
Exercise 5.3 Polynesian
59 The Polynesian languages of the Pacific form a subgroup of the Oceanic branch of the Austronesian family of languages. (1) What are the sound correspondences found in these data? What sound do you reconstruct for the proto-language to represent each sound correspondence set? (2) What sound change or changes have taken place in each of these languages? (3) What is the best reconstruction (proto-form) for 6, 16, 20 and 32? Show how your postulated sound changes apply to each of these to produce the modern forms. NOTE that not all sounds of the proto language are represented in these cognate sets with their sound correspondences. For example, in one not represented clearly here, Tongan has Ø corresponding to l or r of the other languages (reflecting what is usually reconstructed as *r of Proto-Polynesian), distinct from the set in which Tongan has l corresponding to l or r in these sister languages (reflecting Proto- Polynesian *l). This distinction will not be clearly visible in the data presented in this exercise. NOTE: <’> = [ʔ].
Māori Tongan Samoan Rarotongan Hawai’ian gloss 1. tapu tapu tapu tapu kapu ‘forbidden’, ‘taboo’ 2. pito pito — pito piko ‘navel’ 3. puhi puhi — pu’i puhi ‘blow’ 4. taha tafa tafa ta’a kaha ‘side’ ‘edge’ 5. tae ta’e tae tae kae ‘excrement’ ‘trash’ 6. taŋata taŋata taŋata taŋata kanaka ‘man, person’ 7. tai tahi tai tai kai ‘sea’ 8a. kaha kafa ’afa ka’a ’aha ‘strong’ 8b. ma:rohi- ma:lohi ma:losi ma:ro’i — ‘strong’ 9. karo kalo ’alo karo ’alo ‘dodge’ 10. aka aka a’a aka a’a ‘root’ 11. au ’ahu au au au ‘gall’ 12. uru ’ulu ulu uru ulu ‘head’ ‘tip of weapon’ ‘centre’ 13. uhi ufi ufi u’i uhi ‘yam’ 14. ahi afi afi a’i ahi ‘fire’ 15. φa: fa: fa: ’a: ha: ‘four’ 16. φeke feke fe’e ’eke he’e ‘octopus’ 17. ika ika i’a ika i’a ‘fish’ 18. ihu ihu isu (puta-)i’u ihu ‘nose’ ‘nostril’ (puta ‘hole’) 19. hau hau sau ’au hau ‘dew’ ‘wind’ (hauku: ‘dew’ [-ku: ‘showery weather’]) 20. hika — si’a ’ika hi’a ‘firemaking’ 21. hiku hiku si’u ’iku hi’u ‘tail’ ‘fishtail’ 22. ake hake a’e ake a’e ‘up’
60 Māori Tongan Samoan Rarotongan Hawai’ian gloss 23. uru — ulu uru ulu ‘enter’ 24. maŋa maŋa maŋa maŋa mana ‘branch’ 25. mau ma’u mau mau mau ‘constant’ ‘fixed’ 26. mara — mala mara mala ‘fermented food’ ‘marinated’ 27. noho nofo nofo no’o noho ‘sit’ 28. ŋaru ŋalu ŋalu ŋaru nalu ‘wave’ 29. ŋutu ŋutu ŋutu ŋutu nuku ‘mouth’ 30. waka vaka va’a vaka wa’a ‘canoe’ 31. wae va’e vae vae wae ‘leg’ 32. raho laho laso ra’o laho ‘scrotum’ ‘testicle’ 33. rou lohu lou rou lou ‘fruit-picking ‘long forked stick’ pole’ 34. rua ua lua rua lua ‘two’
(1) What are the sound correspondences found in these data? What sound do you reconstruct for the proto-language to represent each sound correspondence set? Sound correspondencess and reconstruction: Māori Tongan Samoan Rarotongan Hawai’ian Proto-Polynesian 1 p p p p p *p In cognate sets 1.-3. 2 t t t t k *t In cognate sets 1., 2., 4.-7., 29. 3 k k ’ k ’ *k In cognate sets 8a.-10., 16.-17., 20.-22., 30. 4 ɸ f f ’ h *f- In cognate sets 15.-16. 5 h f f ’ h *-f- In cognate sets 4., 8a., 13., 14. 6 h h s ’ h *s In cognate sets 3., 8b., 18.-21., 32. 7 m m m m m *m In cognate sets 8b., 24.-26. 8 n n n n n *n In cognate set 27. 9 ŋ ŋ ŋ ŋ n *ŋ In cognate sets 6., 24., 28.-29., 34. 10a r l l r l *l In cognate sets 8b., 9., 12., 23., 26., 28., 32.-33. (10b r Ø l r l *r In cognate set 34.) 11 w v v v w *w In cognate sets 30.-31.
61 Māori Tongan Samoan Rarotongan Hawai’ian Proto-Polynesian 12 Ø ’ Ø Ø Ø *ʔ In cognate sets 5., 25., 31. 13 Ø h Ø Ø Ø *h In cognate sets 7., 11., 33. 14 i i i i i *i In cognate sets 2., 3., 7., etc. 15 a a a a a *a In cognate sets 1., 4., etc. 16 o o o o o *o In cognate sets 2., 8b., 9., etc. 17 u u u u u *u In cognate sets 1., 3., 12., etc.
(2) What sound change or sound changes have taken place in each of these languages?: In Māori, Samoan, Rarotongan, Hawai‘ian 1. *ʔ and *h > Ø /V__ V (in Sound Correspondences [= SC] 12, 13) (ʔ and h were retained only in Samoan)
In Māori 2. *f > h /V __ V (in SC 5)
In Hawai‘ian (and Rarotongan) 3. *f > h (in SC 4, 5)
In Māori, Tongan, Hawai‘ian (and Rarotongan) 4. *s > h (in SC 6)
In Tongan and Hawai‘ian 5. *k > ʔ (in SC 5)
In Māori and Rarotongan 6. *l > r (in SC 10a) Note that with only one cognate set for 10b, in which Tongan has Ø corresponding to l or r in the other languages, there is not much you can say. However if additional data were added, you might be able to formulate better hypotheses about how to explain this. For example, if we add Tongan huu ‘enter’ to cognate set 23., the SC would have to cognate sets exemplifying it. This SC is generally held to represent Proto-Polynesian *r and to be in contrast with that of SC 10a, representing *l. The Proto-Polynesian word in 23. is reconstructed as *huru.
In Tongan, Samoan, Rarotongan 7. *w > v (in SC 11) For the correspondence set reconstructed with *w, it might be suspected that it was originally *v (under majority wins) which changed to w in two of the languages. However, in terms of directionality, w > v is more common than v to w. Actually, it may be that the w had [β] (a bilabial fricative) as an allophone in most of these languages.
62
In Rarotongan 8. h > ʔ (in 4, 5, 6) This assumes that in Rarotongan *f and *s went through an intermediate state in which *f and *s > h, shared with other languages, and then later in its independent history Rarotongan h (from all sources) > ʔ.
In Hawai‘ian: 9. *t > k (in 2) 10. *ŋ > n (in 9)
In Māori 11. *f > ɸ /#__ (in 4) NOTE: the reconstruction of *f fits these data and may be correct, though some might prefer to reconstruct *ɸ (a bilabial rather than a labiodental fricative). Reasons might be that (1) some of the things written as f may in fact actually be bilabial; (2) in directionality ɸ > f may be more expected than f > ɸ; and (3) f in a number of cases is due to later European influence, where European languages with labiodenetal [f] influenced other languages to shift bilabial [ɸ] to [f].
(3) What is the best reconstruction (proto-form) for 6, 16, 20 and 32? Reconstructions: 6. *taŋata ‘man, person’ 16. *feke ‘octopus’ 20. *sika ‘firemaking 32. *laso ‘scrotum, testicle’
Show how your postulated sound changes apply to each of these to produce the modern forms. The following derivations show the application of these sound changes in sequence. When the change does not apply to a particular language, it is left blank; when the sound change takes place in the language but does not apply to the particular form in question, this is signaled by “--”.
6. *taŋata Changes Māori Tongan Samoan Rarotongan Hawai‘ian 1 ------2 -- 3 -- -- 4 ------5 -- -- 6 ------7 ------8 -- 9 kaŋaka 10 kanaka 11 -- Result: taŋata taŋata taŋata taŋata kanaka
63
16. *feke ‘octopus’ Changes Māori Tongan Samoan Rarotongan Hawai‘ian 1 ------2 -- 3 heke heke 4 ------5 fe’e he’e 6 ------7 ------8 ’eke 9 -- 10 -- 11 ɸeke Result: ɸeke feke fe’e ’eke he’e
20. *sika ‘firemaking’ Changes Māori Tongan Samoan Rarotongan Hawai‘ian 1 ------2 -- 3 hika -- -- 4 -- hika hika 5 si’a hi’a 6 -- -- 7 -- -- 8 ’ika 9 -- 10 -- 11 -- Result: hika -- si’a ’ika hi’a
32. *laso ‘scrotum, testicle’ Changes Māori Tongan Samoan Rarotongan Hawai‘ian 1 ------2 -- 3 -- -- 4 laho laho laho laho 5 -- -- 6 raho raho 7 ------8 ra’o 9 -- 10 -- 11 -- Result: raho laho laso ra’o laho
64 Relative Chronology: The exercise does not call for you to work out and state the relative chronology. However, it may be important to point out that change 1. (*ʔ and *h > Ø /V__ V) had to take place prior to 2. (*f > h /V __ V in Māori) and before 3. (*f > h) in Hawai‘ian and Rarotongan, and before 8. (h > ʔ) in Rarotongan, for otherwise the h and ʔ produced by these changes would have gone on in change 1. to be lost in these languages. However, since that did not happen, the loss of h and ʔ by 1. must have taken place before these later cases of h and ʔ were created by subsequent sound changes in these languages.
Exercise 5.4 Orokolo-Toaripi (Eleman languages)
Orokolo and Toaripi are two closely related Eleman languages (usually assigned to the Trans- New Guinea grouping, though this is as yet uncertain). Compare the data presented here and reconstruct Proto-Orokolo-Toaripi. (1) List the sound correspondences you find. (2) Give the proto-sounds you reconstruct to represent these. (3) Present the sound changes which you postulate that each language has undergone. (4) If there is any relative chronology involved among these changes, state what it is and the evidence for it. (5) Give your reconstruction of 12, 25 and 35 together with how the individual sound changes apply to these to produce the modern forms. NOTE: for this problem, consider Orokolo r and l the same sound. Do not struggle over the difference between ae and ai in no. 38.
Toaripi Orokolo gloss 1. uti uki ‘bone’ 2. ete eke ‘vagina’ 3. tete keke ‘fish scales’ 4. tao kao ‘tooth’ 5. toare koare ‘senior’ 6. tola kora ‘tree’ 7. tolotolo korokoro ‘leaves’ 8. tapare kapare ‘grease’ 9. torea korea ‘theft’ 10. turuturu kurukuru ‘thundering’
11. aite aire ‘after’ 12. kite kile ‘mat’ 13. lauta laura ‘flame tree’ 14. ita ila ‘pig’ 15. puta pura ‘cloth’ 16. uta ura ‘hole’
17. fi hi ‘cry’ 18. firu hiru ‘portion’ 19. fe he ‘penis’ 20. fere here ‘betel nut’ 21. fapai hapa ‘open’
65 22. fave have ‘stone’ 23. forerai horera ‘appear’ 24. furi huri ‘pus’ 25. afutae ahurae ‘ashes’
26. sisia hihia ‘sour’ 27. siri hiri ‘mildew’ 28. ase ahe ‘sugarcane’ 29. seseroro heheroro ‘thin’ 30. sare hare ‘sun, day’ 31. sarea harea ‘sorcery’ 32. soa hoa ‘time’ 33. sua hua ‘pigeon’ 34. susu huhu ‘plank’ 35. farisa harita ‘arrow’ 36. marisa marita ‘girl’ 37. taisa kaita ‘paddle’ 38. saesa haita ‘dish’
39.[=12] kite kile ‘mat’ 40. kiva kiva ‘care’ 41. koko koko ‘narrow’ 42. ekaka ekaka ‘fish’
(1) Sound correspondences and (2) reconstructed proto-sounds: Toaripi Orokolo Proto-Eleman t k *t (In 1.-10., 37.) 2 -t- -r- (-l-) *t (In 11.-16., 25.) 3 -s- -t- *s (In 35.-38.) 4 k k *k (In 12., 39.-42.) 5 -r- -r- *r (In 5.-10., 18., 20., 23.-24., 27., 29.-31.,35.-36.) 6 f h *f (In 17.-25., 35.) 7 v v *v (In 22., 40.) 8 s h *s (In 26.-34., 38.) 9 m m *m (In 36.) 10 p p *p (In 8., 15., 21.)
Note that the reconstruction of the correspondence sets involving the sounds t, r, s, and k may appear difficult. In the sets where both languages have the same sound, we reconstruct that particular sound, so *k for sound correspondence [SC] 4 (k : k), and *r intervocalically for SC 5 (-r- : -r-). Given that we reconstruct *k correspondence set 4 (k : k), we must reconstruct something other than *k for SC 1 (t : k), since the sounds of this correspondence set occur in the same environments where SC 4 (k : k) is found. Therefore, we reconstruct *t for 1 (t : k) and postulate a sound change of *t > k in Orokolo. The correspondence sets 2 (-t- : -r-) and 5 (r : r) can occur in the same environments, so they require different reconstructions. Since we already reconstructed *r intervocalically for SC
66 5 (-r- : -r-), SC 2 (-t- : -r-) requires a different explanation. For it, we reconstruct *t and postulate a sound change in Orokolo of *t > r between vowels when the preceding vowel is high and the following vowel is non-high (see charts below). This leave SC 3 (-s- : -t-) unaccounted for. Since *t is taken already by the reconstruction for correspondence sets 1 (t : k) and 2 (-t- : -r-), it is not available for reconstructing SC 3, as SC 2 and SC 3 are found in the same environments. This leaves *s the best hypothesis for 3 (-s- : -t-), even though it requires the postulation of a less likely sound change *s > t intervocalically between i and a in Orokolo, an environment that is different from those of SC 8 (s : h), for which *s is the best reconstruction. While a change of *s > t goes against the typical direction of sound change (t > s is much more common), in its favor is the fact that there seems to be no other more plausible way to account for the contrasting correspondence sets, and for Orokolo now to have no s, with h as the only fricative, is not inconsistent with the phonology of some other languages of Papua New Guinea. To discover the different conditioning environments for the sounds in these correspondence sets, it is helpful to draw up lists or a chart of the environments where the sounds occur, as for example in the following: k : k t : k t : r r : r s : t s : h #__i u__i i__e i__e i__a #__i #__o e__e u__a e__e #__e e__a #__ao i__a o__e #__a a__a #__o e__a #__o #__a u__i #__u #__u i__i i__i #__e e__o e__e u__u o__o u__u a__e a__e a__i
A comparison of the t : k column (SC 1) and the t : r column (SC 2) reveals that while both can occur after high vowels and both can occur before non-high vowels, t : r (SC 2) can occur only when the preceding vowel is high AND the following vowel is non-high. In the case of t : k (SC 1), both the preceding and following vowel is high, or both non-high, or it is in word-initial position, but never in a context between a high vowel and a non-high vowel. While this is unusual conditioning, it does seem to be what accounts for these data, allowing us to reconstruct *t for both sound correspondences.
(3) Sound changes: In Orokolo 1. *t > r /high V__non-high V 2. *t > k 3. *s > t /i__a 4. *f and *s > h (fricative > h) In Toaripi: no changes.
(4) Relative chronology:
67 The chronological order of the changes was: first change 1., then 2., then 3., and finally 4. Change 3. had to precede change 3., for otherwise original *s would have changed to h by change 4. and there would have been no s left in the forms to undergo the change s > t in change 3., giving the wrong results, as for example in 35. *farisa ‘arrow’ > hariha by change 4., so that no s remained to turn to t by change 3., giving erroneous ✘hariha rather than harita in Orokolo, which is the proper outcome which comes about as the result of change 3. preceding change 4. Change 3. took place after chances 1. and 2. had run their course, since otherwise, *s would change to t by 3. and then both this new t and original t would have changed to r (in 1.) or to k (in 2.), giving erroneous results, as in: 36. *marisa ‘girl’ > marita by 3., then marita > ✘marira by 1. In the correct temporal order, 36. *marisa ‘girl’ > marita by 3., after changes 1. and 2. were no longer operable, giving the correct outcome. Changes 1., 2., and 4. do not interact with one another, involving different sounds or different environments, so no relative chronology can be established among them.
(5) Give your reconstruction of 12, 25 and 35 together with how the individual sound changes apply to these to produce the modern forms. In Orokolo: 12. *kite ‘mat’ 25. *afutae ‘ashes’ 35. *farisa ‘arrow’ 1. *t > r/l /high V__non-high V kile afurae -- 2. *t > k ------3. *s > t /i__a -- -- farita 4. *f and *s > h (fricative > h) -- ahurae harita Result: kile ahurae harita In Toaripi: no changes: kite, afutae, farisa
Exercise 5.5 Lencan Compare the cognates from the two Lencan languages (both of which have recently become extinct: Chilanga Lenca was spoken in El Salvador; Honduran Lenca was spoken in Honduras). Work only with the consonants in this problem (the changes involving the vowels are too complex to solve with these data alone). (1) Set up the correspondence sets; (2) reconstruct the sounds of Proto-Lencan; (3) find and list the sound changes which took place in each language; and (4) determine what the relative chronology may have been in any cases where more than one change took place in either individual language, if there is evidence which shows this. NOTE: t’, k’ and ts’ are glottalized consonants. Also, these data do not provide enough information for you to recover all the consonants of the proto-language, so that it will be difficult to apply steps 5 and 6 here.
Honduran Chilanga gloss Lenca 1. pe pe ‘two’ 2. lepa lepa ‘jaguar’ 3. puki puka ‘big’ 4. ta ta ‘cornfield’ 5. tem tem ‘louse’ 6. ke ke ‘stone’
68 7. kuma kumam ‘fingernail, claw’ 8. katu katu ‘spider’ 9. waktik watih ‘sandals’ 10. kakma k’ama ‘gourd’ 11. siksik sisih ‘shrimp’ 12. nek neh ‘tooth’ 13. insek ints’eh ‘beak’ 14. taw t’aw ‘house’ 15. tutu t’ut’u ‘flea’ 16. kin k’in ‘road’ 17. kunan k’ula ‘who’ 18. kelkin k’elkin ‘tortilla griddle’ 19. sewe ts’ewe ‘monkey’ 20. saj ts’aj ‘five’ 21. musu muts’u ‘liver’ 22. sak- ts’ih- ‘to wash’ 23. lawa lawa ‘three’ 24. liwa- liwa- ‘to buy’ 25. tal- tal- ‘to drink’ 26. wala wala ‘raccoon’ 27. was wal ‘water’ 28. asa alah ‘head’ 29. wasan wila ‘urine’ 30. wara wara ‘river’ 31. siri sirih ‘star’ 32. sili sili ‘iron tree’ (tree species) 33. suri-sur ʃurih ‘squirrel’ [NOTE: suri-sur involves reduplications; just compare the suri- segment of it] 34. saj- ʃej- ‘to want’ 35. so ʃo ‘rain’ 36. suna ʃila ‘flower’ 37. soko ʃoko ‘white’ 38. sak ʃah ‘firewood’ 39. wewe wewe ‘baby’ 40. jet- jete- ‘to laugh’ 41. juku juku ‘coyol palm’ (palm tree species) 42. sa ʃam ‘good’
(1) Set up the correspondence sets; (2) reconstruct the sounds of Proto-Lencan: Honduran Chilanga Proto-Lencan Lenca (1) -k- Ø *k *k > Ø /__C in Chilanga In 9. watktik / watih, 10. kakma / k’ama 11. siksik, sisih
(2) -k -h *k *k > h /__# in Chilanga
69 In 9. watktik / watih, 11. siksik / sisih, 12.nek / neh, 13. insek / ints’eh, 22. sak- / ts’ih-, 38. sak / ʃah
(3) k k *k (elsewhere) in both languages In 3. puki / puka, 18. kelkin / k’ilkin
(4) -h -Ø *h *h > Ø /__# in Hondoruan Lenca In 28. asa / alah, 31. siri / sirih, 33. suri(-sur) / ʃurih
(5) -n -Ø *n *n > Ø /a__# in Chilanga In 17. kunan / k’ula, 29. wasan, wila
(6) n l *n *n > l /V__V in Chilanga In 17. kunan / k’ula, 36. suna / ʃila (A possible alternative reconstruction: n : l *l with *l > n /u__ in Honduran Lenca; there is, however, no good phonetic reason why u should condition l to change to n).
(7) n n *n (elsewhere) in both languages In 12. nek / neh, 13. insek / ints’eh, 16. kin / k’in
(8) l l *l In 2. lepa / lepa, 22. lawa / lawa, 24. liwa- / liwa-, 25. tal- / tal-, 26. wala / wala, 32. sili / sili, 18. kelkin / k’elkin, 31. sili / sili
(9) s l *ɫ *ɫ > s in Honduran Lenca (ɫ = voiceless “l”) *ɫ > l in Chilanga In 27. was / wal, 28. asa / alah, 29. wasan / wila
(10) s ts’ *ts’ *ts’ > s in Honduran Lenca In 13. insek / ints’eh, 19. sewe / ts’ewe, 20. saj / ts’aj, 21. musu / muts’u, 22. sak- / ts’ih-
(11) t t’ *t’ *t’ > t in Honduran Lenca In 14. taw / t’aw, 15. tutu / t’ut’u
(12) k k’ *k *k’ > k in Honduran Lenca In 10. kakma / k’ama, 16. kin / k’in, 17. kunan / k’ula, 18. kelkin / k’elkin
(13) s s *s In 11. siksik / sisih, 31. siri / sirih, 32. sili / sili (14) s ʃ *ʃ *ʃ > s in Honduran Lenca In 33. suri(-sur) / ʃurih, 34. saj- / ʃej-, 35. so / ʃo, 36. suna / ʃila, 37. soko / ʃoko, 38. sak / ʃah, 42. sa / ʃam
(3) Sound changes in each language (C = Chilanga, HL = Honduran Lenca): In Chilanga: (C1a) *k > Ø /__C, (C1b) *k > h /__# It is possible to combine (C1a) and (C1b) into a single change, *k changing in different, complementary environments; *k was lost before a consonant, and changed to h word- finally.
70 (C2a) *n > Ø /a__#, (C2b) *n > l /V__V It is possible to combine (C2a) and (C2b) into a single change, *n changing in different, complementary environments; *n was lost after a at the end of words, and changed to l between vowels. (C3) *ɫ > l (voiceless “l” > plain “l”) In Honduran Lenca: (HL1) *h > Ø /__# Final h was lost. (HL2) *ɫ > s (voiceless “l” > s) (HL3) C’ > C (glottalized C > plain counterpart, *t’ > t, *k’ > k, *ts’ > ts) Glottalized consonants lost their glottalization. [NOTE: this is true of these data as given, though it is equally plausible that Honduran Lenca did retain glottalized consonants, but that they were just not recorded in the attestations available to us.] (HL4) ts (from *ts’ by HL3) > s Plain ts, after (HL3) changed *ts’ to ts, went on to change to s (ts’ > ts > s) (HL5) *ʃ > s
(4) Determine what the relative chronology may have been: There is no basis for determining relative chronology for any of these changes in either language, since the sounds involved in one change do not figure in any other change. Only in the case of (HL3) and (HL4) do they interact, so that in (HL3) *ts’ > ts produces ts, which is an intermediate stage that then goes on, changing ts > s in (HL4). However, while it is phonetically likely that there was an intermediate stage with ts, this is not necessary. It could be assumed that (HL4) was rather of the form *ts’ > s (directly or possibly through an intermediate stage of ts) but that this happened before (HL3) took place. It is also possible to assume that perhaps (HL3) involved only the stops and that *ts’ was not part of that change, making (HL3) and (HL4) independent of one other, with no interaction.
Exercise: 5.6. Uto-Aztecan
(1) State the sound correspondences. (2) Present the sound that you reconstruct for Proto-Uto- Aztecan for each sound correspondence. (3) List the sound changes that you observe in the various languages. Ignore vowel length and do not attempt to reconstruct the vowels for this exercise. There are not sufficient examples in the data given here to be able to reconstruct the full set of Proto-Uto-Aztecan sounds. Assume that correspondences found in only a single cognate set would recur if more data were present. Attempt to reconstruct the consonants in initial and medial position and state the sound changes you postulate to get from your reconstructed consonants to the actual forms in the various daughter languages. Given the paucity of forms cited here, you may have to postulate some sound changes on the basis of poor evidence, in hopes of confirming (or disconfirming) them when more data are brought into the picture. It is often the case that the historical linguist must work with incomplete or imperfect data, so the challenge here of attempting to reconstruct with less than complete information is a realistic experience. (NOTE: a few examples have been regularized, slightly modified, in order to avoid complications for the reconstruction.) Tohono
71 Cupeño Hopi Comache O’odham Huichol Nahuatl 1 paqa pa:qa- paka ‘arrow’ wa:-pka haka a:ka ‘reed’ 2 pa- pa:hɨ pa: wa- ha: a: ‘water’ 3 pah pa:y- pahi- -- hai- e:y ‘three’ 4 puš po:si pui wuhi hɨši i:š ‘eye’ 5 pi- pöhɨ puʔe wo:g hu:ye: oʔ ‘road’ 6 -- pɨtɨ pɨhtɨ we:č he:te ete- ‘heavy’ 7 tama tama ta:ma ta:-tami tame tla:n ‘tooth’ 8 -- teni -- čini teni te:n ‘mouth’ 9 tuu-, tula tö:vɨ (kuh-)tu:bi ču:d tɨ: ti:l ‘charcoal’ 10 -- tös- tus(oyuni) čuhi tɨsi tiš ‘grind, flour’ ‘grindstone’ 11 -- qa:si -- kahio -- ikši ‘leg, thigh’ 12 qwaše kwasi kwasɨ-pɨ bahi kwaši ikwši ‘cooked, ripe’ 13 qwaš -- kwasi bahi kwaši -- ‘tail’ 14 qweʔ- kwɨʔɨ -- baʔa -kwaʔa kwa ‘eat’ 15 -- kwita kwita- bi:t kwita kwitla ‘excrement’ 16 maqa maqa maka ma:k -- maka ‘give’ 17 -- mo:ki -- mu:ki mɨki miki ‘kill, die’ 18 mala mata -- maččud ma:ta: matla ‘grindstone’ 19 naqa na:qa na:ki na:k naka nakas ‘ear’ 20 nema nɨ:ma nɨ:ma nem nema -- ‘liver’ 21 waxe la:ki -- gaki -waki wa:ki ‘dry’ 22 wexi- lökö woko -- huku oko ‘pine’ (Based on Stubbs 2011.)
(3) List the sound changes that you observe in the various languages.
(1) State the sound correspondences. (2) Present the sound that you reconstruct for the Proto- Uto-Aztecan for each sound correspondence:
Initial: Tohono Cupeño Hopi Comanche O’odham Huichol Nahuatl PUA Cognate set number p p p w h Ø *p 1, 2, 3, 4, 5, 6 t t t t t tl *t 7 t t (t) č t t *t 8, 9, 10 -- q - k k k *k 11 qw kw kw b kw kw *kw 12, 13, 14, 15 m m m m m m *m 16, 17, 18 n n n n n n *n 19, 20 w l g w w w *w 21 w l g w h Ø *w 22
72
Medial: Tohono Cupeño Hopi Comanche O’odham Huichol Nahuatl PUA Cognate set number l t t č t t *t 6 l t t č t tl *t 18 -- t t t t tl *t 15 q q k k k k *k 1, 16, 19 x k k k k k *k 21, 22 š s Ø h š š *s 4 š s s h š š *s 12, 13 m m m m m -n *m 7, 20 -- n -- n n n *n 8 ʔ ʔ -- ʔ ʔ Ø *ʔ 14
(3) Sound changes: In Cupeño: (C1) *k > q /__ a, *kw > qw /__ a (in 12, 13, 14) Velars (k, kw) became uvulars (q, qw) before a. (C2) *k > x /__ i, e (in 21, 22) (C3) *t > l /V__V (intervocalic t > l) (C4) *s > š /__ i, e (4, 12, 13) In Hopi: (H1) *k > q /__ a (1, 11, 16, 19) (H2) *w > l /#__ (in 21, 21) Initial w became l. In Tohono O’odham: (TO1) *w > g (in 21) (TO2) *t > č /__high V (as in 6, 8, 9, 10) (TO3) *kw > b (as in 12, 13, 14, 15) (TO4) *s > h /V__V (intervocalic s > h, as in 4, 12, 13) (TO5) *p > w in Tohono O’odham (as in 1, 2, 4, 5, 6) In Huichol and Nahuatl: (HN1) *p > h [NOTE: if more data were given, it would be seen that this change affected only initial *p.] In Nahuatl: (N1) h (from *p) > Ø If we assume that Huichol and Nahuatl belong to the same subgroup, more closely related to one another than to other languages, we may postulate (HN1) *p > h in the parent of the subgroup to which Huichol and Nahuatl belong, then later Nahuatl h > Ø in (N1). (N1) *t > tl /__a Here, *t became the lateral affricate tl before a (as in 6, 7, 15, 18). (N2) *s > š /__ i (as in 4, 10, 11, 12) (N3) *ʔ > Ø (or h, ʔ > Ø, see N1 above, reflexes of *p) (N4) *a > Ø /Voiceless Stop __ Voiceless Stop Here, a was lost between voiceless stops (as in 11, 12)
73 (N5) Ø > i /#__CC Here, i was epenthesized at the beginning of a word before a consonant cluster (in 11, 12).
Relative chronology (not called for in the instructions): There are a few cases of obvious relative chronology. For example, a must have been lost in (N4) in order for the consonant cluster to come into existence which is the conditioning environment for (N5) in Nahuatl: in 11 *kasi ‘leg, thigh’ > ksi (by N5), then > iksi (by N6), and to ikši (by N2) – (N2) could have taken place in any temporal order with respect to (N4) and (N5). In Tohono O’odham, (TO1) *w > g must have preceded (TO5) *p > w, for otherwise the w from original *p (by TO5) and original *w would have both ended up as g; original *pa: ‘water’ would have become wa: by (TO5) and then gone on erroneously to ✘ga: by (TO1), if that had been the temporal sequence. Therefore, (TO1) took place first, changing *w to g, and then (TO5) later changed *p to w (but this new w did not change further, as it would have if the order of the changes had been reversed).
Exercise 5.7 Jicaquean
Jicaquean is a family of two languages in Honduras. Jicaque (Jicaque of El Palmar) is extinct; Tol (Jicaque of La Montaña de la Flor) is still spoken by a few hundred people, but has become extinct or nearly so everywhere except in the village of La Montaña de la Flor. Reconstruct Proto-Jicaquean; state the sound correspondences which you encounter in the following cognate sets, and reconstruct a proto-sound for each. State the sound changes that have taken place in each language. HINT: your reconstruction should include the following sounds: p t ts k ʔ i _ u ph th tsh kh e o p’ t’ ts’ k’ a s l m n w j h What happens to each of the proto-sounds which you reconstruct in initial and in final position in these two languages? Can you make guesses about an appropriate reconstruction and sound changes to account for sounds in medial positions? NOTE: the correspondences involving affricates and sibilants are quite complex, and you will need to pay special attention to the possibilities for combining some of the initial correspondence sets with some of the medial ones as reflecting the same proto-sound. The consonants p’, t’, ts’, k’ are glottalized. The accent mark on a vowel (for example á) means that it is stressed; this is not relevant to the sound changes. In a few cases, a non-initial h does not match well in the two languages; ignore this, since it is due to changes for which you do not have enough evidence in these data. The hyphen (-) before some words, as in 9 (-rɨk), means that these occur with some other morpheme before them which is not relevant and so is not presented here.
74 Jicaque Tol gloss 1. pe pe ‘stone’ 2. pit pis ‘meat’ 3. pɨné pɨné ‘big’ 4. pɨga- pɨʔa- ‘jaguar’ 5. pen pel ‘flea’ 6. kamba kampa ‘far, long’ 7. arba- alpa ‘above’ 8. to-bwe to-pwe ‘to burn’ 9. -rɨk -lɨp ‘lip’ 10. kek kep ‘woman’ 11. ik hip ‘you’ 12. huruk hulup ‘grain’ (of corn) 13. huk hup ‘he, that’ 14. nak nap ‘I’ 15. -kuk -kup ‘we’ 16. te te ‘black’ 17. tek tek ‘leg’ 18. tebé tepé ‘he died’ 19. tɨt tɨt’ ‘louse’ 20. mandɨ mantɨ ‘vulture’ 21. n-gon n-kol ‘my belly’ 22. harek halek ‘arrow’ 23. mak mak ‘foreigner’ 24. n-abuk n-ajphuk ‘my head’ 25. kon kom ‘liver’ 26.[=6] kamba kampa ‘far, long’ 27. pɨrɨk pɨlɨk ‘much’ 28. keré kelé ‘nephew’ 29. mik mik ‘nose’ 30. korok kolok ‘spider’ 31. phe phe ‘white’ 32. phen phel ‘arm, shoulder’ 33. -pha -pha ‘dry’ 34. phɨja phɨja ‘tobacco’ 35. m-bat m-phats’ ‘my ear’ 36. lɨbɨ- lɨphɨ ‘wind’ 37. phɨbɨh phɨphɨh ‘ashes’ 38. urubana (j)uluphana ‘four’ 39. ten them ‘boa constrictor’ 40. tut thuth ‘spit’ 41. peten pethel ‘wasp’ 42. kun khul ‘fish’ 43. ke-ke (kh)ekhe ‘agouti’ [NOTE: keke is a reduplicated form and should be treated as the root ke- repeated, rather than as having an intervocalic -k-]
75 44. kan khan ‘bed’ 45. kere khhele ‘bone’ 46. to-gon- to-khol ‘to grind’ 47. kujuh khujuh ‘parrot’ 48. pɨt p’ɨs ‘deer’ 49. m-bɨj m-p’ɨj ‘my body’ 50. pɨčá p’ɨsá ‘macaw’ 51. -te -t’e ‘to cut’ 52.[=19] tɨt tɨt’ ‘louse’ 53. -tja -t’ja ‘to be late’ 54. mata mat’a ‘two’ 55. kat ʔas ‘blood’ 56. kot ʔos ‘I sit, am’ 57. kaw- ʔaw-a ‘fire’ 58. kona ʔona ‘sour’ 59. kan ʔan ‘zapote’ (fruit) 60.[=4] pɨga- pɨʔa- ‘jaguar’ 61. te-ga te-ʔa ‘to give’ 62. čok sok’ ‘tail’ 63. čorin tsolin ‘salt’ 64. ču(h) tsu ‘blue’ 65. čiwiri -tsiwil- ‘to lie’ 66. čigin- tsikin ‘summer’ 67. čoʔ- tsoʔ- ‘to nurse’ 68. čuba tsupa ‘to tie’ 69. nočot notsots ‘fly’ 70. ʃeme tsheme ‘horn’ 71. ʃijó tshijó ‘dog’ 72. ʃe(w) tshew ‘scorpion’ 73. čin ts’il ‘hair, root’ 74. -čun ts’ul ‘intestines’ 75. čoron ts’olol ‘oak’ 76. čih ts’ih- ‘caterpillar’ 77. te-neče te-nets’e ‘to sing’ 78. ločak lots’ak ‘sun’ 79. m-bat m-phats’ ‘my ear’ 80. čot sots’ ‘owl’ 81. -čɨ -sɨ ‘water’ 82. čok sok’ ‘tail’ 83. [=2] pit pis ‘meat’ 84. -mut mus ‘smoke’ 85. hoč(uruk) hos- ‘his heart’ 86.[=50] pɨčá p’ɨsá ‘macaw’ 87. mon mol ‘cloud’ 88.[=25] kon kom ‘liver’ 89. ma ma ‘land’
76 90. wa wa ‘house’ 91. wara wala ‘forehead’ 92. jo jo ‘tree’ 93. he he ‘red’ 94.[=22] harek halek ‘arrow’ (Data from Campbell and Oltrogge 1980)
State the sound correspondences which you encounter in the cognate sets, and reconstruct a proto-sound for each. Jicaque Tol Proto-Jicaque 1 p p *p (In 1.-5., 27., 41., 60., 83.) 2 b p *p (In 6.-8., 18., 26., 68.) (/m__, V__V/w) 3 t t *t (In 8., 16.-19., 46., 52., 61., 77.) 4 d t *t (In 20.) 5 k k *k (In 6., 10., 15., 17., 22.-30., 78., 88., 94.) 6 -g- -k- *k (In 66.) 7 -k -p *p (In 9.-15.) 8 ph- ph- *ph (In 31.-34.) 9 -b- ph- *ph (In 24., 35.-38.) 10 t- th- *th (In 39.-41.) 11 k kh *kh (In 43.-45., 47.) 12 -g- kh *kh (In 46.) 13 p p’ *p’ (In 48., 50., 86.) 14 -b- -p’- *p’ (In 49.) (/m__) 15 t t’ *t’ (In 19., 51.-54.) 16 -k’ -k’ *k’ (In 62., 82.) 17 k- ʔ- *k’ (In 55.-59. 18 -g- -ʔ- *k’ (In 4., 60.-61.) 19 č ts *ts (In 63.-69.) 20 -t -ts *ts (In 69.) 21 ʃ- tsh *tsh (In 70.-72.) 22 č ts’ *ts’ (In 73.-78.) 23 -t ts’ *ts’ (In 35., 79.-80. 24 č s *s (In 62., 80.-82., 85.-86.) 25 -t -s *s (In 2., 48., 55.-56., 83.-84.) 26 l- l- *l (In 36., 78.) 27 -r- -l- *l (In 7., 9., 12., 22., 27.,-28., 30., 38., 45., 63., 65., 75., 91., 94.) 28 -n -l *l (In 5., 21., 32., 41.-42., 46., 73.-75., 87.) 29 m m *m (In 6., 20., 23., 26., 29., 35., 54., 70., 79., 84., 87., 89.) 30 n n *n (In 3., 14., 20., 38., 44., 58.-59., 63., 66., 69., 77.) 31 -n -m *m (In 25., 39., 88.) 32 w w *w (In 8., 57., 72., 90.-91.) 33 j j *j (In 34., 47., 49., 53., 71., 92.) 34 h h *h (In 12.-13., 22., 37., 47., 85., 93.-94.) 35 ʔ ʔ *ʔ (In 67.) Vowel correspondences are the same in the two language and so the numbers that identify
77 the cognate sets exemplifying the vowel correspondences are not listed here: Jicaque Tol Proto-Jicaque 36 i i *i 37 e e *e 38 a a *a 39 o o *o 40 u u *u 41 ɨ ɨ *ɨ
Sound changes in Jicaque: 1. *p’, *t’, *k’, *ts’ > p, t, k, ts respectively Glottalized stops and affricates lost glottalization, changing to their plain counterparts. (In sound correspondences [SC] 13-18, 22.-23.) Some of these sounds changed further in later changes. [NOTE: it is quite possible that Jicaque maintained glottalized consonants but that the attestations we have simply did not record them. However, the fact that they become voiced in change 3. suggests that they did lose glottalization, at least in some environments.] 2. *p > k /__ # (In SC 7) 3. Stop > voiced /voiced C __ and /V__ Stops are voiced following a voiced consonant or a vowel (p, t, k > b, d, g /voiced C __, and p, t, k > b, d, g /V __). This includes plain and aspirated voiceless stops; it also includes original glottalized stops which lost glottalization in sound change 1. (In SCs 2, 4, 6, 9, 12, 14, 18.) 4. ts, *s > t /__ # Sibilants (ts, s) > t word-finally. This includes also ts from original *ts’ which lost glottalization in sound change 1. (In SCs 20, 23, 25.) 5. ts > č This includes also ts originally from *ts’ which lost glottalization in sound change 1. (In SC 19.) (Changes 4. and 5. could perhaps be combined into a single change by notational conventions, since they apply in complementary environments, cognates involved in 4. word-finally, and those in SC 5. elsewhere, not word-finally.) 6a. *l > n /__# 6b.*l > r /V__Segment Here, *l > n word-finally (in 6a) (in SC 28), and *l > r (in 6b) when between a vowel and another segment (in SC 27). It is possible to combine 6a. and 6b. into a single change using appropriate notation, since they involve complementary environments. (In 6b. the l must be followed by another segment, a vowel in most instances, a consonant in some (as for example in cognate set 7.), since without mentioning a following segment, if *l followed a vowel but were in final position, it would become n. It is of course possible to conclude that these changes took place in sequence, first 6a., so that once *l became n, there would be no remaining instances of final l and so it would not be necessary to specify a following segment in the environment for 6b., just *l > r /V__.
Sound changes in Tol:
78 7. *k’ > ʔ /__V Here, *k’ > ʔ before a vowel, that is, initially and intervocalically, where a vowel follows in both. (In SCs 17., 18.)
Exercise 5.8 K’ichean
K’ichean is a subgroup of the Mayan family. Compare these cognate forms and set up the sound correspondences; propose the most appropriate reconstruction for the sound in the proto- language for each, and write the sound changes which account for the developments in the daughter languages. Are any instances found in any of the individual languages in which it is necessary to state what the relative chronology of changes was? NOTE: ɓ = voiced imploded bilabial stop; t’, ts’, č’, k’, q’, m’, w’ = glottalized consonants. In Uspanteko, the accent mark over the vowel, as in ò:x ‘avocado’, indicates falling tone. Although the correspondence set in which Q’eqchi’ h corresponds to x of the other languages is not found in these data before u, ignore this – this sound correspondence occurs in general with no restrictions that have anything to do with u. NOTE: y = IPA [j], š = IPA [ʃ], č = IPA [tʃ] C’ = glottalized [ejective] consonants.
Abbreviations: Kak = Kaqchikel Tz’u = Tz’utujil K’i = K’iche’ Poq = Popomam Usp = Uspanteko Q’eq = Q’eqchi’ PK = Proto-K’ichean
Kak Tz’u K’i Poq Usp Q’eq gloss 1. pak pak pak pak pak pak ‘custard apple’ 2. pur pur pur pur pur pur ‘snail’ 3. pim pim pim pim pim pim ‘thick’ 4. toʔ toʔ toʔ toʔ toʔ toʔ ‘to help’ 5. tox tox tox tox tox tox ‘to pay’ 6. kiʔ kiʔ kiʔ kiʔ kiʔ kiʔ ‘sweet’ 7. ka:ʔ ka:ʔ ka:ʔ ka:ʔ ka:ʔ ka:ʔ ‘quern’(metate) 8. k’el k’el k’el k’el k’el (k’el) ‘parrot’ 9. qa- qa- qa- qa- qa- qa- ‘our’ 10. qul qul qul — qul — ‘neck’ 11. q’o:l q’ol q’o:l q’o:l q’o:l q’o:l ‘resin, pitch’ 12. q’an q’an q’an q’an q’an q’an ‘yellow’ 13. si:p si:p si:p si:p si:p si:p ‘tick’ 14. saq saq saq saq saq saq ‘white’ 15. tsuy tsuy tsuh suh tsuh suh ‘water gourd’ 16. uts uts uts us uts us ‘good’ 17. tsats tsats tsats sas tsats sas ‘thick’
79 18. ts’iʔ ts’iʔ ts’iʔ ts’iʔ ts’iʔ ‘ts’iʔ’ ‘dog’ 19. če:ʔ če:ʔ če:ʔ če:ʔ če:ʔ če:ʔ ‘tree, wood’ Kak Tz’u K’i Poq Usp Q’eq gloss 20. ču:n ču:n ču:n ču:n ču:n ču:n ‘lime’ 21. č’o:p č’o:p č’o:p č’o:p č’o:p č’o:p ‘pineapple’ 22. xul xul xul xul xul xul ‘hole, cave’ 23. winaq winaq winaq winaq winaq kwinq ‘person’ 24. we:š we:š we:š we:š — kwe:š ‘trousers’ 25. ya:x ya:x ya:x ya:x ya:x ya:x ‘genitals, shame’ 26. mu:x mu:x mu:x mu:x mù:x mu:h ‘shade’ 27. o:x o:x o:x o:x ò:x o:h ‘avocado’ 28. ča:x ča:x ča:x ča:x čà:x ča:h ‘ashes’ 29. tu:x tu:x tu:x tu:x tù:x tu:h ‘steambath’ 30. q’i:x q’i:x q’i:x q’i:x q’ì:x (-q’ih) ‘day, sun’ 31. ka:x ka:x ka:x ka:x kà:x — ‘sky’ 32. čax čax čax čax čax čax ‘pine’ 33. k’ax k’ax k’ax k’ax k’ax k’ax ‘flour’ 34. k’o:x k’o:x k’o:x k’o:x k’o:x k’o:x ‘mask’ 35. ɓa:y ɓa:y ɓa:h w’a:h ɓa:h ɓa:h ‘gopher’ 36. ɓa:q ɓa:q ɓa:q w’a:q ɓaq ɓaq ‘bone’ 37. ɓe:y ɓe:y ɓe:h w’e:h ɓe:h ɓe:h ‘road’ 38. siɓ siɓ siɓ sim’ siɓ siɓ ‘smoke’ 39. xaɓ xaɓ xaɓ xam’ xaɓ haɓ ‘rain’ 40. xuku:ʔ xuku:ʔ xuku:ɓ xuku:m’ xuku:ɓ xukuɓ ‘canoe, trough’ 41. a:q’aʔ a:q’aʔ a:q’aɓ a:q’am’ a:q’aɓ (a:q’ɓ) ‘night’ 42. xal xal xal xal xal hal ‘ear of corn’ 43. xe:y xe:y xe:h xe:h xe:h he:h ‘tail’ 44. č’o:y č’o:y č’o:h č’o:h č’o:h č’o:h ‘mouse, rat’ 45. k’yaq k’yaq k’yaq k’aq k’aq k’aq ‘flea’ 46. kyaq kyaq kyaq kaq kaq kaq ‘red’ 47. (i)kyaq’ (i)kyaq’ kyaq’ kaq’ — — ‘guava’ 48. išk’yaq šk’yaq išk’yaq išk’aq išk’aq — ‘fingernail’ 49. winaq winaq winaq winaq winaq kwinq ‘person’ 50. šikin šikin šikin šikin šikin (šikn) ‘ear’ 51. išoq išoq išoq išoq — išq ‘woman’ 52. nimaq nimaq nimaq nimaq nimaq ninq ‘big’ (plural) 53. sanik sanik sanik (sanik) sanik sank ‘ant’ 54. suʔt suʔt suʔt suʔt sù:t’ (suʔut) ‘cloth, kerchief’ 55. poʔt poʔt poʔt poʔt pò:t’ poʔot ‘blouse’ 56. piʔq piʔq piʔq piʔq pì:q’ — ‘corncob’ 57. atiʔt atiʔt atiʔt atiʔt atì:t’ atiʔt ‘grandmother’ 58. k’ax k’ax k’ax k’ax k’ax k’ax ‘flour’ 59. k’ay k’ay k’ah k’ah k’ah k’ah ‘bitter’ 60. k’ay k’ay k’ay k’ay k’ay k’ay ‘to sell’ 61. mo:y mo:y mo:y mo:y mo:y mo:y ‘blind’ (dark) 62. ča:x ča:x ča:x ča:x čà:x ca:h ‘ashes’
80 63. čax čax čax čax čax čax ‘pine’ 64. č’ax č’ax č’ax — č’ax č’ax ‘to wash’ 65. č’ay č’ay č’ay č’ay — — ‘to hit’
Sound correspondences, and reconstruction for the sound in the proto-language: Kak Tz’u K’i Poq Usp Q’eq PK [1] x x x x x x *x [2] x x x x x h *xy [3] y y h h h h *h [4] ts ts s s ts ts *ts [5] ɓ- ɓ- ɓ- w’- ɓ- ɓ- *ɓ [6] -ɓ -ɓ -ɓ -m’ -ɓ -ɓ *ɓ [7] -ʔ -ʔ -ɓ -m’ -ɓ -ɓ *ɓ [8] k k k k k k *k [9] ky ky ky k k k *k [10] VʔC VʔC VʔC VʔC V̀:C VʔC *VʔC (C= stop) [11] V: V: V: V: V̀: V: *V: /__xy [12] V V V V V Ø *V /VC__C# [13] m m m m m n *m /__q [14] w w w w w kw *w /#__ [15] p p p p p p *p [16] t t t t t t *t [18] č č č č č č * č [17] k k k k k k *k [18] q q q q q q *q [19] ʔ ʔ ʔ ʔ ʔ ʔ *ʔ [20] ts’ ts’ ts’ ts’ ts’ ts’ *ts’ [21] č’ č’ č’ č’ č’ č’ * č’ [22] k’ k’ k’ k’ k’ k’ *k’ [23] q’ q’ q’ q’ q’ q’ *q’ [24] s s s s s s *s [25] š š š š š š *š [25] l l l l l l *l [26] r r r r r r *r [27] m m m m m m *m [28] n n n n n n *n [29] y y y y y y *y
The vowels have identical correspondences across the languages (with the exception of the change affecting tone in Uspantec, and the deletion of short vowels and the shortening of long vowels in final syllables followed by a final consonant in Q’eqchi’. The vowels are thus reconstructed as: *i *u *e *o *a Plus their long counterparts.
81
Sound changes in the daughter languages: (1) Short V > Ø /VC__C# In Q’eqchi’ in SC [12] (as in examples 23. *winaq > kwinq, 52. *nimaq > ninq, 54. *sanik > sank) Short vowels were deleted in the final syllable before a final consonant. (2) *m > n /__q In Q’eqchi’ in SC [13] (as in 52. *nimaq > ninq) (3) *ɓ > ʔ /VC__# In Kaqchikel and Tz’utujil in SC [7] (as in 40. *xukuɓ, 41. *a:q’aɓ) Final imploded “b” changed to glottal stop. (4a) *ɓ > w’ /#__ In Poqoman in SC [5] (as in 35. *ɓah, 36. *ɓaq, 37. *ɓe:h) Initial imploded “b” changed to glottalized “w”. (4b) *ɓ > m’ /__# In Poqomam in SC [6] (as in 38. *siɓ, 39. *xaɓ, 40. *xukuɓ, 41. a:q’aɓ) Final imploded “b” changed to glottalized “m”. Here, (4a) and (4b) can be combined into a single change, since they affect complementary environments; imploded “b” became a glottalized w word-initially and a glottalized m word finally. [NOTE: with more data, it would become clear that the change was to m’ word-finally and to w’ in all other positions, not just initially.] (5) Ø > y /k(’)__Vq(’) In Kaqchikel, Tz’utujil, K’iche’, Poqomam in SC [9] (as in 45. *k’aq > k’yaq, 46. *kaq > kyaq, 47. *ikaq’ > (i)kyaq’, 48. *išk’aq > išk’yaq). A y was inserted after a velar stop (k or k’) when followed by a vowel and a uvular stop (q or q’). [This seems to be dissimilation, making a verlar more distinct from a uvular within the same syllable.] (6) VʔC > V̀:C’ In Uspanteko (where the C = stop) in SC [10] (as in 54. *suʔt > sù:t’, 55. *poʔt > pò:t/, 56. *piʔq > pì:q’, 57. *atiʔt > atì:t’). The sequence of a Vowel, glottal stop (ʔ), and a stop together resulted in the stop becoming glottalized, the glottal stop disappearing, and the preceding vowel being lengthened with falling tone. (7) V: > V̀: /__ *xy In Uspanteko in SC [11] (as in 26. *mu:xy > mù:x, 27. *o:xy > ò:x, 28. [also 62.] *ča:xy > čà:x, 29. *tu:xy > tù:x, 30. *q’i:xy > q’ì:x, 31. *ka:xy > *kà:x). A long vowel developed falling tone before the fronted velar fricative *xy. (8) *xy > h In Q’eqchi’ in SC [2] (as in 26., 27., 28. [= 62.], 29., 30., 39., 42., 43.) (9) *xy > x In Kaqchikel, Tz’utujil, K’iche’, Poqomam, and Uspanteko in SC [2] as (in 26., 27., 28. [= 62.], 29., 30., 39., 42., 43.). (10) *ts > s In Poqomam and Q’eqchi’ in SC [4] (as in 15. *tsuh > suh, 16. *uts > us, 17. *tsats > sas). (11) *h > y /__# In Kaqchikel and Tz’utujil in SC [3] (as in 15., 35., 43., 59.). Kaqchikel and Tz’utujil share the change of final h to y. (12) *w > kw /#__ In Q’eqchi’ in SC [14] (as in 23., 24.). In Q’eqchi’, initial *w was strengthened to kw. [In these data there is no indication that *w did not become kw in all contexts; had more data been provided, it would become clear that syllable final w remained unchanged.]
Relative chronology: Change (1) took place before (2) in Q’eqchi’. In 52. *nimaq > nimq by change (1) comes before nimq > ninq by change (2). If (2) had occurred before (1), then (2) would not apply to *nimaq; it could not change m to n before q, since there would be no q following m, and if (1) then took place after (2), we would end up with ✘nimq, the wrong outcome.
82 Change (7) took place before change (9) in Uspanteko. For example, in 28. (same as 62.) *ča:xy > čà:xy ‘ashes’ by (7), and then čà:xy > čà:x by (9), but in 63. ‘pine’ change (7) does not apply to *ča:x, since there is no xy in the environment to condition the tone change in the preceding vowel, and change (9) does not apply, again, since there was no xy to change to x: *ča:xy ‘ashes’ *ča:x ‘pine’ Change (7) čà:xy -- Change (9) čà:x -- Result čà:x ča:x However, if change (9) had taken place earlier than change (7) in the history of Uspanteko, then in 28./62. ‘ashes’, xy > x (*ča:xy > *ča:x) by (9), and if (7) had come later, there would no longer have been the xy needed for (7) to lower the tone on the preceding vowel, giving the wrong form: *ča:xy ‘ashes’ *ča:x ‘pine’ Change (9) ča:x -- Change (7) -- -- Result ✘ča:x ča:x
Exercise 5.9 Quechuan
Quechuan is a family of several languages spoken in the Andes region of South America, with varieties found in Columbia, Ecuador, Peru, Bolivia and Argentina. Compare the cognates from the languages listed here. Set up the correspondence sets; reconstruct the sounds of Proto- Quechuan; find and list the sound changes which took place in each language (variety); determine what the relative chronology may have been in any cases where more than one change took place in an individual language (variety), if there is evidence which shows this. What do you think the inventory of Proto-Quechuan sounds was? (Note that there is some controversy about the historical status of glottalized consonants (p’, t’, č’, k’, q’) and aspirated consonants (ph, th, čh, kh, qh) in Quechuan. For the purposes of this exercise do not try to reconstruct them, but rather treat those few which occur (in the Cuzco variety) as though they were equal to the plain counterparts.) (NOTE: [Ŋ̣] = uvular nasal; y = IPA [j]; l y = IPA [lj]; ny = IPA [nj]. Abbreviations: An = Ancash Ju = Junín Ca = Cajamarca Am = Amazonas Ec = Ecuador Ay = Ayacucho Cu = Cuzco
An Ju Ca Am Ec Ay Cu 1. paka- paka- paka- paka- paka- paka- paka- ‘begin’ 2. apa- apa- apa- apa- apa- apa- apa- ‘wash’ 3. rapra lapla rapra rapra — rapra raφra ‘leaf, wing’ 4. pampa pampa pamba pamba pamba pampa pampa ‘plains’
83 An Ju Ca Am Ec Ay Cu 5. tapu- tapu- tapu- tapu- tapu- tapu- tapu- ‘ask’ 6. wata- wata- wata- wata- wata- wata- wata- ‘tie’ 7. utka utka utka utka — utka uskha ‘cotton’ 8. inti inti indi indi indi inti inti ‘sun’ 9. kimsa kimsa kimsa kimsa kimsa kimsa kimsa ‘three’ 10. puka puka — puka puka puka puka ‘red’ 11. haksa- saksa- saksa- saxsa- saxsa- saksa- saxsa- ‘be full, fed up’ 12. kuŋka kuŋka kuŋga kuŋga kuŋga kuŋka kuŋka neck 13. qam am qam kam kaŋ χam qaŋ ‘you’ (SG) 14. qoha usa qosa kusa kusa χosa qosa ‘husband’ 15. waɢa- waʔa- waga- waka- waka- waχa- waqa- ‘cry’ 16. hoχta suʔta soχta sukta suxta soχta soχta ‘six’ 17. heŊ̣ɢa siŋʔa seŊ̣ɢa siŋga siŋga seŊ̣χa seŊ̣qa ‘nose’ 18. tsaki čaki čaki čaki čaki čaki č’aki ‘dry’
19. mutsa- muča- muča- muča- muča- muča- muč’a- ‘kiss’ 20. mantsa- manča- manča- manča- manča- manča- manča- ‘fear, be afraid’ 21. putska- pučka- pučka- pučka- puʃka- pučka- puska- ‘to thread’ 22. e:tsa ayča ayča e:ča ayča ayča ayča ‘meat’
23. čaki č̣aki č̣aki č̣aki č̣aki čaki čaki ‘foot’ 24. kača- kač̣a- kač̣a- kač̣a- kač̣a- kača- kača- ‘send’ 25. učpa uč̣pa uč̣pa uč̣pa uʃpa učpa uspha ‘ashes’ 26. kički kič̣ki kič̣ki kič̣ki kiʃki kički k’iski ‘narrow’
27. haru- salu- saru- saru- saru- saru- saru- ‘to step on’ 28. hara sala sara sara sara sara sara ‘maize, corn’ 29. qaha asa qasa kasa kasa χasa qasa ‘ice’ 30. isqoŋ isʔuŋ esqoŋ iʃkuŋ iʃkuŋ isχoŋ esqoŋ ‘nine’ 31. — aysa- aysa- e:sa- aysa- aysa- aysa- ‘pull’ 32. waʃa waʃa waʃa waʃa waʃa wasa wasa ‘behind’ 33. iʃke: iʃkay iʃkay iʃke: iʃkay iskay iskay ‘two’
34. hatuŋ hatuŋ atuŋ atuŋ hatuŋ hatuŋ hatuŋ ‘big’ 35. hutsa huča uča uča huča huča huča ‘fault’ 36. humpi humpi — umbi humbi humpi hump’i ‘sweat’
37. laki lyaki ʒaki ǰaki ʒaki lyaki lyaki ‘pain, trouble’ 38. kila kilya kiʒa kiǰa kiʒa kilya kilya ‘moon’ 39. alba alypa aʃpa aǰpa aʒpa alypa halyp’a ‘land’ 40. aylu aylyu ayʒu e: ǰu ayʒu aylyu aylyu ‘family’
41. rima- lima- rima- rima- rima- rima- rima- ‘to speak’ 42. karu kalu karu karu karu karu karu ‘far’
84 An Ju Ca Am Ec Ay Cu 43. warmi walmi warmi warmi warmi warmi warmi ‘woman’ 44. waχra waʔla waχra wakra — waχra waχra ‘horn’ 45. nina nina nina nina nina nina nina ‘fire’ 46. yana yana yana yana yana yana yana ‘black’ 47. wayna wayna wayna wayna wayna wayna wayna ‘young man’ 48. aŋya- aŋya- aŋya- aŋya- aŋya- aŋya- aŋya- ‘to reprove’
49. nawi nyawi nyawi nyawi nyawi nyawi nyawi ‘eye’ 50. wanu- wanyu- wanyu- wanyu- wanyu- wanyu- wanyu- ‘to die’
51. qepa ipa qepa kipa kipa χepa qhepa ‘behind’ 52. weɢe wiʔi — wiki wiki weχe weqe ‘tear(drop)’(NOUN)
53. qeʃpi- iʃpi- — kiʃpi- kiʃpi- χespi- qespi- ‘to escape’ 54. qo- u- qo- — ku- χo- qo- ‘to give’ 55. qoŊ̣ga- uŋʔa- qoŊ̣ga- kuŋga- kuŋga- χoŊ̣χa- qoŊ̣qa- ‘to forget’
Set up the correspondence sets, reconstruct the sounds of Proto-Quechuan: An Ju Ca Am Ec Ay Cu Proto-Quechua 1 p p p p p p p *p (In 1., 2., 3., 5., 10., 21., 25., 51., 53.) 2 p p p p p p ɸ *p (In 3.) 3 p p b b b p p *p (In 36.) 4 b p p p p p p *p (In 40.) 5 t t t t t t t *t (In 5.-6., 16., 34. ) 6 t t t t t t s *t (In 7.) 7 t t d d d t t *t (In 8.) 8 k k k k k k k *k (In 9.-10., 12., 21., 23.-24., 26., 33., 37.-38., 42.) 9 k k k x x k x *k (In 11.) 10 k k g g g k k *k (In 12. ) 11 q Ø q k k χ q *q (In 13.-14., 51., 53.-55.) An Ju Ca Am Ec Ay Cu Proto-Quechua 12 ɢ ʔ ɢ k k χ q *q (In 15., 52.) 13 ɢ ʔ ɢ g g χ q *q (In 17., 55.) 14 χ ʔ χ k x χ χ *q (In 16., 44.) 15 ts č č č č č č *č (In 18.-20., 22.) 16 ts č č č ʃ č s *č (In 21.) 17 č č̣ č̣ č̣ č č č *č (In 23-24.) 18 č č̣ č̣ č̣ ʃ č s *č̣ (In 25.-26.) 19 h s s s s s s *s (In 27.-29., 31.) 20 s s s s s s s *s (In 9., 11., 31.) 21 s s s ʃ ʃ s s *s (In 30.) 22 ʃ ʃ ʃ ʃ ʃ s s *s (In 32., 33., 53.) 23 l ly ʒ ǰ ʒ ly ly * ly (In 37.-38., 40.)
85 24 l ly ʃ ǰ ʒ ly ly * ly (In 39.) 25 r l r r r r r *r (In 41.-44.) 26 m m m m m m m *m (In 4., 9., 19.-20., 36., 41., 43.) 27 m m m m ŋ m ŋ *m (In 13.) 28 n n n n n n n *n (In 20., 45.-47.) 29 ŋ ŋ ŋ ŋ ŋ ŋ ŋ *n (In 12.) 30 ŋ̣ ŋ ŋ̣ ŋ ŋ ŋ̣ ŋ̣ *n (In 17., 30., 55.) 31 n ny ny ny ny ny ny *ny (In 49.-50.) 32 w w w w w w w *w (In 43.-44., 47., 49.-50., 52.) 33 y y y y y y y *y (In 46.-48.) 34 h h Ø Ø h h h *h (In 34.-36.) 35 a a a a a a a *a (In many.) 36 i i i i i i i *i (In many.) 37 e i e i i e e *i (In 17., 51.-53.) 38 i i e i i i e *i (In 30.) 39 u u u u u u u *u (In many.) 40 o u o u u o o *u (In. 14., 16., 30., 54.-55.) 41 e: ay ay e: ay ay ay *ay (In. 22., 31., 33., 39.) 42 ay ay ay ay ay ay ay *ay (In 40., 47.)
Sound changes: find and list the sound changes which took place in each language (variety):
Changes affecting all varieties, as allophones in Proto-Quechua: Qu 1. Nasal assimilation: *n > ŋ /__ velar (In sound correspondence [SC] 29), *n > Ŋ̣ /__ uvular (In SC 30), *n > m /__ labial (In SCs 4, 38), and *n remains n before alveolars and alveopalatals (In cognate sets 8., 20.). Qu 2. *n > ŋ /__ # Final *n became ŋ (an allophone of *n). (In SC 30). Qu 3. High vowel > non-high /q __ and /__ (n)q (In SCx 37-40) *i > e, and *u > o in the environment of q, either after q or before *q or *nq. Some languages later underwent additional minor changes in these vowels resulting in some variation in the individual languages, where the environment of the vowel lowering change has been extended to include *sq. The languages which lost the q : k contrast (Junín, Amazonas, and Ecuador) also do not have e or o in the environment of earlier *q. The vowel lowering was allophonic, not phonemic, as e and o did not contrast with i and u, but rather were automatic co-articulation affects as a consequence of the uvular consonant in the environment. Since the lowering was tied so closely to the presence of a uvular, when q > k in these languages, there was no longer a uvular to condition these vowels to lower, and as a consequence, the languages lacking q also lack e and o as allophones of /i/ and /u/.
Ancash sound changes: An 1. *p > b /__l (In SC 4) An 2a. *q > ɢ /V__V (In SG 12) An 2b. *q > ɢ /Nasal __ (In SC 13)
86 In Ancash the uvular stop *q became voiced (ɢ) both between vowels and after nasals. An 3. *q > χ /__C (In SC 14) An 4. *č > ts (In SCs 15, 16) An 5. *č̣ > č (In SC 17) An 6. *s > h /#__ a (In SC 19) An 7. Palatalized sonorants > non-palatalized *ly > l (In SCs 23, 24), *ny > n (In SC 31) An 8. *ay > e: /__ obstruent Here, ay monophongized to e: before obstruents (as in SC 41 [examples in cognate sets 22., 31., 33., 39.]), but not before sonorants (l and n, as in SC 42 [examples in cognate sets 40, 47.]).
Ancash relative chronology: An 4. (*č > ts) preceded An 5. (*č̣ > č), for otherwise *č̣ would have become č first (by An 5), and then the new č and original *č would both have become ts (by An 4), giving wrong results (as in SC 17, which retained č). This is, in effect, a small chain shift; it would be considered a pull chain if first *č > ts, leaving an opening for *č̣ to move into the space vacated by former č. It would be a push chain if evidence could be found that *č̣ started moving towards the space of č, causing č to move to ts in order to maintain the contrast and avoid merger with the new č from earlier *č̣.
Junín sound changes: Ju 1. *q > Ø /#__ (In SC 11) Initial *q was lost. Ju 2. *q > ʔ /V__V (In SCs 12, 13) *q became a glottal stop between vowels. Ju 3. *s > ʃ /__ C (In SC 22) (perhaps only before velars and uvulars). (Ju lost allophonic e and o as a consequence of the loss of the contrastive uvulars.)
Cajamarca sound changes: Ca 1. Stop > voiced /__ Nasal (In SCs 3, 7, 10, 13) Here, p, t, k, q > b, d, g, ɢ, respectively, before a nasal. Ca 2. *q > χ /__ C (In SC 14) Ca 3. *ly > ʒ /__ V (In SC 23), *ly > ʃ /__ C (In 24) Ca 4. i > e /__ Cq (In SC 38, also in SC 37) It is presumed that Cajamarca extended the environment of the Qu 3. allophonic change to applying not just when adjacent to q or before *nq, but to include also any consonant followed by q (before *sq or *nq, the only clusters involving q as the second member in these data).
Amazonas sound changes: Am 1. k > x /__ C (In SC 9) Am 2. *q > k (In SCs 11, 12, 13, 14) Am 3. Stop > voiced /__ Nasal (In SCs 3, 7, 10, 13) p, t, k > b, d, g, respectively, before nasals. Am 4. *s > ʃ /__ C, (In SC 21) *s > ʃ /V__V (In SC 22)
87 *s > ʃ before consonants and intervocalically, but not initially nor after consonants (as in cognate set 9.). Am 5. *ly > ǰ (In SCs 23, 24) Am 6. *ay > e: /__ obstruent Here, ay monophongized to e: before obstruents (as in SC 41 [examples in cognate sets 22., 31., 33., 39.]), but not before sonorants (l and n, as in 42 [examples in cognate sets 40, 47.]). (Am lost allophonic e and o as a consequence of the loss of *q.)
Amazonas relative chronology: Am 1. must have preceded Am. 2., since otherwise original *q > k (by Am 2) and then both original *k and the new k from *q would have become x (by Am 1), as in cognate set 44. *waqra > wakra by Am 2. then ✘waxra by Am 1., an error. With Am. 1. taking place before Am 2., the outcome gives the right result: to *waqra Am. 1. did not apply, as if affects only k not q, then later Am 2. applied to give the correct wakra.
Ecuador sound changes: Ec 1. *q > k (In SCs 11, 12, 13, 14) Uvulars and velars merged to velars. Ec 2. k > x /__ C (In SCs 9, 14) Both original *k and new k from q in Ec. 1 become x before a following consonant. Ec 3. Stop > voiced /__ Nasal (In SCs 3, 7, 10, 13) p, t, k > b, d, g, respectively, before nasals. Ec 4. *s > ʃ /__ C, (In SC 21), *s > ʃ /V__V (In SC 22) Here, *s > ʃ before consonants and intervocalically, but not initially nor after consonants (as in cognate set 9.). Ec 5. *č > ʃ /__ C (In SC 16) Ec 6. *ly > ʒ (In SCs 23, 24) (Ec lost allophonic e and o as a consequence of the loss of *q.)
Ecuador Relative Chronology: Ec 1. preceded Ec. 2., since original *q > k and then both original *k and the new k from *q became x, as in cognate set 44. *waqra > wakra by Ec 1., then waxra by Am 1.
Ayacucho sound changes: Ay 1. *q > χ (In SCs 11, 12, 13, 14)
Cuzco sound changes: Cu 1. *č̣ > *č (In SCs 15-18) *č and *č̣ merged to č. Cu 2. Stops and affricates > fricatives /__ C (In SCs 3, 6, 9, 14, 16, 18) Before a consonant: p > ɸ, t > s, č (from both *č and *č̣) > s, k > x, q > χ. Presumably in the change of č > s there was an intermediate stage: č > ʃ > s, which would require another change, Cu 2a. ʃ > s. Cu 3. i > e /__ Cq (In SC 38, also SC 37)
88 It is presumed that Cuzco extended the environment of the Qu 3. allophonic change to apply not just being adjacent to q or before *nq, but to include also any consonant followed by q (before *sq or *nq, the only clusters involving q as the second member in these data).
Proto-Quechua phonemic inventory, as reflected in these data: p t *č *č̣ k q s h r ly m n ny w y
i u a
89
EXERCISES FOR CHAPTER 6 LINGUISTICS CLASSIFICATION
6.1 Exercise: Polynesian subgrouping
There are numerous Polynesian languages, of which the ones in this exercise are only a sample. Polynesian is a branch (subgroup) of the Austronesian family. Consider the two cognate sets from these five Polynesian languages. What sound change has taken place? Observe which languages underwent the change and which ones did not, and on this basis draw all the alternative possible family trees that could account for the subgrouping of these languages. (Clue: which languages share an innovation? Which languages share a retention?) Note that the medial consonant, spelled
Proto-Polynesian *songi ‘smell’ *sae ‘to tear’ Tongan hongi hae Niuean hongi hee Samoan sogi sae East Futuna sogi sae Luangiua songi sae (Data from Otsuka 2005.)
The sound change is: *s > h The languages which underwent this change are: Tongan and Niuean. Tongan and Niuean share the innovation (*s > h). Samoan, East Futuma, and Luangiua share only a retention (of *s).
Possible family trees based on the shared innovation: so long as Tongan and Niuean are joined together in a single branch, all other logically possible combinations are possible family trees, based on these data. Keep in mind, however, various family trees remain possibilities because the evidence is insufficient to rule them out, and these are not subgroupings that are supported by shared innovations – only the branch containing both Tongan and Niuean is securely supported by the shared innovation (and even then, we are assuming that the change did not take place independently in the two languages, although that possibility cannot be ruled out). The possible family trees are:
(1) Proto-Polynesian
Tongan Niuean Samoan East Futuna Luangiua
90 (Four independent branches, one of which is a subgroup of the two languages Tongan and Niuean.)
(2) Proto-Polynesian
Tongan Niuean Samoan East Futuna Luangiua (Two branches, one subgroup with Tongan and Niuean, another subgroup with the other three languages, Samoan, East Futuna, and Luangiua. There is evidence, the shared innovation, for the subgroup that includes Tongan and Niuean; there is no evidence to confirm or deny the other logically possible subgroup, joining Samoan, East Futuna, and Luangiwa.)
(3) Proto-Polynesian
Tongan Niuean Samoan East Futuna Luangiua (Three branches, one with Tongan and Niuean, one with Samoan and East Futuna, and one with independent Luangiua. There is evidence, the shared innovation, for the subgroup that includes Tongan and Niuean; there is no evidence to confirm or deny the other possible subgroup that joining Samoan and East Futuna, nor the independence of Luangiwa.)
(4) Proto-Polynesian
Tongan Niuean Samoa East Futuna Luangiua (Three branches, one subgroup with Tongan and Niuean, one with independent Samoan, and one with East Futuna and Luangiua.)
Some other possibilities would be: (1) A subgroup of Tongan-Niuean-Samoan, with East Futuna and Luangiua as two independence branches, where Tongan-Niuean and Samoan would then later split, and after the
91 Samoan split off, the ancestor of Tongan-Niuean underwent the change that came to be reflected as a shared innovation in Tongan and Samoan. (2) As in (1), with a subgroup of Tongan-Niuean-Samoan but with East Futuna and Luangiua together in a subgroup, though there is no evidence of a Futuna-Luagiua subgroup in these data. (3) A subgroup of Tongan-Niuean-East Futuna, with Samoan and Luangiua as separate branches. (4) As in (3), with a subgroup of Tongan-Niuean-East Futuna, but with Samoan and Luangiua together in a subgroup, though there is no evidence of a Samoan-Luagiua subgroup in these data.
6.2 Exercise: Barbacoan subgrouping
Barbacoan is a small family of languages spoken in Colombia and Ecuador; its members include: Awa Pit, Cha’palaachi, Guambiano, Totoró, and Tsafiqui. Consider the following sound changes in the various Barbacoan languages. On the basis of these, draw the most likely family tree which represents the subgrouping of the Barbacoan family. (Note, č = IPA tʃ, š = IPA ʃ.)
1. Proto-Barbacoan *t > č / __ i (when before i) in Guambiano 2. Proto-Barbacoan *t > č / __ i (when before i) in Totoró 3. Proto-Barbacoan *t > š / __ i (when before i) in Awa Pit 4. Proto-Barbacoan *t > tʂ elsewhere (when not before i) in Guambiano 5. Proto-Barbacoan *t > tʂ elsewhere (when not before i) in Totoró 6. Proto-Barbacoan *t > s elsewhere (when not before i) in Awa Pit 7. Proto-Barbacoan *š > Ø /__ # (lost word-finally) in Guambiano 8. Proto-Barbacoan *š > Ø /__ # (lost word-finally) in Totoró 9. Proto-Barbacoan *l > n / #__ i (before i) in Guambiano 10. Proto-Barbacoan *l > n / #__ i (before i) in Totoró 11. Proto-Barbacoan *l > n / #__ i (before i) in Awa Pit 12. Proto-Barbacoan *p, *t, *k > Ø/__ # (stops lost word-finally) in Cha’palaachi 13. Proto-Barbacoan *p, *t, *k > Ø/__ # (stops lost word-finally) in Tsafiqui 14. Proto-Barbacoan *s > h in Cha’palaachi 15. Proto-Barbacoan *s > h in Tsafiqui (Source Curnow and Liddicoat 1998.)
Shared sound changes and generalized treatment of the changes: (1) *t > č /__ i shared by Guambiano, Totoró, and Awa Pit (sound changes 1.-3.) (2) č [from *t by (1)] > š in Awa Pit (in sound change 3., better seen as *t > č > š) (3) *t > tʂ shared by Guambiano, Totoró, and Awa Pit (sound changes 4.-6.) (4) tʂ [from *t by (3)] > s in Awa Pit (in sound change 6., better seen as *t > tʂ > s) (5) *š > Ø /__ # shared by Cha’palaachi and Tsafiqui (sound changes 7.-8.) (6) *l > n /#__ i shared by Guambiano, Totoró, and Awa Pit (sound changes 9.-11.) (7) *p, *t, *k > Ø/__ # shared by Cha’palaachi and Tsafiqui (sound changes 12.-13.) (8) *s > h shared by Cha’palaachi and Tsafiqui (sound changes 14.-5.)
92 Guambiano, Totoró, and Awa Pit share the changes described in (1), (3), and (6). Cha’palaachi and Tsafiqui share those in (5), (7), and (8).
The possible trees must group Guambiano, Totoró, and Awa Pit in one subgroup (based on the shared innovations (1), (3), and (6)), and Cha’palaachi and Tsafiqui in another subgroup (based on the shared innovations (5), (7), and (8)). However, the evidence here is insufficient to determine which of other possible groupings may be more likely (more historically accurate). The possible trees are:
(1) Proto-Barbacoan
Guambiano Totoró Awa Pit Cha’palaachi Tsafiqui (Two subgroups, one with two languages, Cha’palaachi and Tsafiqui, and one with three equally distant languages, Guambiano, Totoró, and Awa Pit.)
(2) Proto-Barbacoan
Guambiano Totoró Awa Pit Cha’palaachi Tsafiqui (Two subgroups, one with two languages, Cha’palaachi and Tsafiqui; the other which is further subdivided into two branches of its own, one with Guambiano and Totoró more closely related to one another, and with Awa Pit more distantly related to these two within this subgroup.)
(3) Proto-Barbacoan
Guambiano Totoró Awa Pit Cha’palaachi Tsafiqui (Two subgroups, one with two languages, Cha’palaachi and Tsafiqui; the other further subdivided into two branches of its own, one branch with Guambiano, and the other with Totoró and Awa Pit more closely related to one another.)
93
EXERCISES FOR CHAPTER 8 INTERNAL RECONSTRUCTION
Exercise 8.1 German internal reconstruction
Compare the following German words; find the variants of forms of the roots (do not be concerned with the forms of the suffixes), and apply internal reconstruction to these. Reconstruct a single original form for the morphemes which have alternate forms, and postulate the changes which you think took place to produce the modern variants. Present your reasoning; why did you choose this solution? (Hint: the criterion of predictability is important in this case.) (German traditional orthography is given in parentheses after the forms, which are presented in phonemic transcription. The ‘e’ of the final syllable in these forms is phonetically closer to [ə] in most dialects, though this is not a relevant fact for solving this problem.)
1. ty:p (Typ) ‘type’ ty:pen (Typen) ‘types’ 2. to:t (tot) ‘dead’ to:te (Tote) ‘dead people’ 3. lak (Lack) ‘varnish’ lake (Lacke) ‘kinds of varnish’ 4. tawp (taub) ‘deaf’ tawbe (Taube) ‘deaf people’ 5. to:t (Tod) ‘death’ to:de (Tode) ‘deaths’ 6. ta:k (Tag) ‘day’ ta:ge (Tage) ‘days’
Reconstruct a single original form for the morphemes: Pre-German root 1. ty:p (Typ) ‘type’ ty:p-en (Typen) ‘types’ *ty:p 2. to:t (tot) ‘dead’ to:t-e (Tote) ‘dead people’ *to:t 3. lak (Lack) ‘varnish’ lak-e (Lacke) ‘kinds of varnish’ *lak 4. taup (taub) ‘deaf’ taub-e (Taube) ‘deaf people’ *taub 5. to:t (Tod) ‘death’ to:d-e (Tode) ‘deaths’ *to:d 6. ta:k (Tag) ‘day’ ta:g-e (Tage) ‘days’ *ta:g Change 1.: stop > voiceless /__# Stops in word-final position become voiceless (as in 4., 5. 6.).
Examples of reconstruction and application of Change 1: 1. *ty:p *ty:p-en 4. *tawb *tawb-e ‘type’ ‘types’ ‘deaf’ ‘deaf people’ Change 1.: -- -- tawp -- Result: ty:p ty:pen tawp tawbe
Reasoning: A proposed alternative reconstruction with voiceless final consonants as original and a change to voice them intervocalically produces erroneous forms; it fails on the criterion of predictability: Change A.: C > voiced /V__V 1. *ty:p *ty:p-en 4. *tawp *tawp-e Change A.: -- ty:ben -- taube Result ty:p ✘ty:ben tawp taube
94 Change A. cannot distinguish between cases such as 4. where the stop ends up voiced between vowels (taube) and cases such as 1. where the change predicts the stop should also be voiced in ty:pen but it is not voiced. The change would produce erroneous ✘ty:ben. Since this proposed alternative cannot accurately predict all the resulting forms, it must be abandoned. The solution with Change 1. devoicing final stops does correctly predict the outcomes of all these forms.
Exercise 8.2 Kaqchikel internal reconstruction
Kaqchikel is a Mayan language of Guatemala. Compare the following words; find the forms which have variants; apply internal reconstruction to these forms. Reconstruct a single original form for the morphemes which have alternate forms, and postulate the changes that you think must have taken place to produce these variants. Present your reasoning; why did you choose this solution and reject other possible hypotheses? (Note that -ir is the inchoative suffix, meaning ‘to become/turn into’, and-isax is the causative suffix.)
1. nax ‘far’ naxt-ir-isax ‘to distance’ 2. čox ‘straight’ čoxm-ir ‘to become straight’ 3. war ‘sleep’ wart-isax ‘to put to sleep’ 4. ax ‘ear of corn’ axn-i ‘of corn’ (i ‘adjective suffix’)
Reconstruct a single original form for the morphemes which have alternate forms: Pre-Kaqchikel root 1. nax ‘far’ naxt-ir-isax ‘to distance’ *naxt 2. čox ‘straight’ čoxm-ir ‘to become straight’ *čoxm 3. war ‘sleep’ wart-isax ‘to put to sleep’ *wart 4. ax ‘ear of corn’ axn-i ‘of corn’ (i ‘adjective suffix’) *axn
Postulate the changes: Change 1.: C > Ø /C __ # The last consonant of a word-final consonant cluster was deleted.
Examples of reconstruction and application of the change: 1. *naxt *naxt-ir-isax 2. *čoxm *čoxm-ir ‘far’ ‘to distance’ ‘straight’ ‘to become straight’ Change 1.: nax -- čox -- Result: nax naxtirisax čox čoxmir
Reasoning: An alternative analysis which would reconstruct the roots with only a single final consonant (*nax, *čox, *war, *ax) and then attempt to predict which consonant was added when not word final, that is, when followed by a suffix, does not work, since there is no logical phonetic basis for predicting which consonant would be added, why a -t- in 1., but -m- in 2., and so on.
95 Exercise 8.3 Sanskrit internal reconstruction
Compare the following forms from Sanskrit. Identify the variants of the various roots and attempt to reconstruct a Pre-Sanskrit form for each root. Note that the reconstructions for the forms in 10–16 are not straightforward and may require some creative thinking on your part. What change do you think took place to produce these forms? Why did you choose this particular analysis and not some other? NOTE: j = [ǰ], IPA [dʒ]; consonants with dots underneath are retroflex.)
Nominative Instrumental 1. šarat ‘autumn’ šarad-ā ‘by autumn’ 2. sampat ‘wealth’ sampad-ā ‘by wealth’ 3. vipat ‘calamity’ vipad-ā ‘by calamity’ 4. marut ‘wind’ marut-ā ‘by wind’ 5. sarit ‘river’ sarit-ā ‘by river’ 6. jagat ‘world’ jagat-ā ‘by world’
Nominative Ablative 7. suhṛt ‘friend’ suhṛd-ā ‘from friend’ 8. sukṛt ‘good deed’ sukṛt-ā ‘from good deed’ 9. sat ‘being’ sat-ā ‘from being’ 10. bhiṣak ‘physician’ bhiṣaj-ā ‘from physician’ 11. ṛtvik ‘priest’ ṛtvij-ā ‘from priest’ 12. yuk ‘yoke’ yuj-ā ‘from yoke’ 13. srak ‘garland’ sraj-ā ‘from garland’ 14. rāṭ ‘king’ rāj-ā ‘from king’ 15. iṭ ‘worship’ ij-ā ‘from worship’ 16. sṛṭ ‘creation’ sṛj-ā ‘from creation’ (Bhat 2001: 33, 91,94)
Variants of the roots and Pre-Sanskrit reconstructions: Pre-Sanskrit 1. šarat šarad- *šarad ‘autumn’ 2. sampat sampad- *sampad ‘wealth’ 3. vipat vipad- *vipad ‘calamity’ 4. marut *marut ‘wind’ 5. sarit *sarit ‘river’ 6. jagat *jagat ‘world’ 7. suhṛt suhṛd- *suhṛd ‘friend’ 8. sukṛt *sukṛt ‘good deed’ 9. sat *sat ‘being’ 10. bhiṣak bhiṣaj- *bhiṣag ‘physician’ 11. ṛtvik ṛtvij- *ṛtvig ‘priest’ 12. yuk yuj- *yug ‘yoke’ 13. srak sraj- *srag ‘garland’ 14. rāṭ rāj- *rāḍ ‘king’ 15. iṭ ij- *iḍ ‘worship’ 16. sṛṭ sṛj- *sṛḍ ‘creation’
96
Change 1. Final devoicing: stop > voiceless /__# Change 2. Palatalization: g, ḍ > j /__ ā *šarad *šarad-a *yug *yug-ā *rāḍ *rāḍ-ā Ch1: šarat -- yuk -- rāṭ -- Ch 2: ------yuj-ā -- rāj-ā Result: šarat šarada yuk yujā rāṭ rājā
Change 2 Palatalization of g and ḍ to ǰ would not be unusual, except here the change takes place before ā, not before front vowels, as is common of palatalization changes cross- linguistically. Given the limitations of the data in the exercise, you might say that the change took place before any suffix vowel: g, ḍ > j /__+V, though if more data were given, we would find out that this change did not happen before just any vowel at the beginning of a suffix. Historically, Sanskrit e > a, and it can be assumed that at an earlier time it was e and ē which conditioned the palatalization in Change 2, like similar changes in many languages before front vowels, and then later e and ē changed to a and ā, respectively, after the preceding g and ḍ had become palatalized. It is unlikely that most people would posit an earlier stage with e and ē for these vowels in order to have a plausible environment for the palatalization in Change 2 and then later merge e and ē with a and ā, respectively, after palatalization had changed the forms. If you did postulate such a solution, it would be within the bounds of internal reconstruction, though hard to justify on the basis of these data alone.
Exercise 8.4 Internal reconstruction of Finnish vowels
Compare the following words; what happens when the i ‘plural’ or i ‘past tense’ morphemes are added to these roots? State what the variants (allomorphs) of the roots are; apply internal reconstruction to these forms. Reconstruct a single original form for each root morpheme and postulate the changes which you think must have taken place to produce these variants. Present your reasoning; why did you choose this solution and reject other possible hypotheses? NOTE: double vowels, such as aa, yy, and so on, are phonetically long vowels ([a:], [y:], etc. Finnish ä = IPA [æ], ö = [ø]. HINT: Native Finnish words do not have (surface) oo, ee, or öö [øø]; rather, Finnish has uo, ie, and yö [yø] where long mid vowels would be expected. The correct answer for words containing these diphthongs does NOT involve the first vowel being lost when i is added (that is, NOT suo + i > soi by loss of u).
1. saa ‘gets’ sai ‘got’ 2. maa ‘land’ mai- ‘lands’ 3. puu ‘tree’ pui- ‘trees’ 4. luu ‘bone’ lui- ‘bones’ 5. pii ‘tooth (of rake)’ pii- ‘teeth’ 6. pää ‘head’ päi- ‘heads’ 7. pyy ‘wood grouse’ pyi- ‘wood grouses’ 8. täi ‘louse’ täi- ‘lice’ 9. suo ‘grants’ soi ‘granted’ 10. suo ‘swamp’ soi- ‘swamps’
97 11. luo ‘creates’ loi- ‘created’ 12. syö ‘eats’ söi ‘ate’ 13. lyö ‘hits’ löi ‘hit’ 14. tie ‘road’ tei- ‘roads’ 15. vie ‘takes’ vei ‘took’ 16. talo ‘house’ taloi- ‘houses’ 17. hillo ‘jam’ hilloi- ‘jams’ 18. halu ‘desire’ halui- ‘desires’ 19. hylly ‘shelf’ hyllyi- ‘shelves’ 20. nukke ‘doll’ nukkei- ‘dolls’ 21. tekö ‘deed’ teköi- ‘deeds’ 22. sata ‘hundred’ satoi- ‘hundreds’ 23. pala ‘piece’ paloi- ‘pieces’ 24. hella ‘stove’(‘cooker’) helloi- ‘stoves’ (‘cookers’) 25. hilkka ‘hood’ hilkkoi- ‘hoods’ 26. hiha ‘sleeve’ hihoi- ‘sleeves’ 27. sota ‘war’ sotei- ‘wars’ (soti- in Modern Finnish) 28. pora ‘drill’ porei- ‘drills’ (pori- in Modern Finnish) 29. muna ‘egg’ munei- ‘eggs’ (muni- in Modern Finnish) 30. rulla ‘roll’ rullei- ‘rolls’ (rulli- in Modern Finnish) 31. tupa ‘cabin’ tupei- ‘cabins’ (tupi- in Modern Finnish) 32. jyvä ‘grain’ jyvei- ‘grains’ (jyvi- in Modern Finnish) 33. hätä ‘distress’ hätei- ‘distresses’ (häti- in Modern Finnish) 34. mökä ‘hullabaloo’ mökei- ‘hullabaloos’ (möki- in Modern Finnish)
Reconstruct a single original form for each root morpheme: Variants Pre-Finnish Pre-Finnish 1. saa / sa- *saa ‘get’ *saa-i- ‘got’ 2. maa / ma- *maa ‘land’ *maa-i- ‘lands’ 3. puu / pu- *puu ‘tree’ *puu-i- ‘trees’ 4. luu / lu- *luu ‘bone’ *luu-i- ‘bones’ 5. pii / pi- *pii- ‘tooth’ *pii-i- ‘teeth (of rake)’ 6. pää / pä- *pää ‘head’ *pää-i- ‘heads’ 7. pyy / py- *pyy ‘wood grouse’ *pyy-i- ‘wood grouses’ 8. täi / tä- * täi ‘louse’ *täi-i- ‘lice’ 9. suo / so- *soo ‘grants’ *soo-i- ‘granted’ 10. suo / so- *soo ‘swamp’ *soo-i- ‘swamps’ 11. luo / lo- *loo ‘creates’ *loo-i- ‘created’ 12. syö / sö- *söö ‘eats’ *söö-i- ‘ate’ 13. lyö /lö- *löö ‘hits’ *löö-i- ‘hit’ 14. tie / te- *tee ‘road’ *tee-i- ‘roads’ 15. vie / ve- *vee ‘takes’ *vee-i- ‘took’ 16. talo *talo ‘house’ *talo-i- ‘houses’ 17. hillo *hillo ‘jam’ *hillo-i- ‘jams’ 18. halu *halu ‘desire’ *halu-i- ‘desires’ 19. hylly *hylly ‘shelf’ *hylly-i- ‘shelves’
98 20. nukke *nukke ‘doll’ *nukke-i- ‘dolls’ 21. hölmö *hölmö ‘fool’ *hölmö-i- ‘fools’ 22. sata / sato- *sata ‘hundred’ *sata-i- ‘hundreds’ 23. pala / palo- *pala ‘piece’ *pala-i- ‘pieces’ 24. hella / hello- *hella ‘stove’ *hella-i- ‘stoves’ (‘cookers’) 25. hilkka / hilkko- *hilkka ‘hood’ *hilkka-i- ‘hoods’ 26. hiha / hiho- *hiha ‘sleeve’ *hiha-i- ‘sleeves’ 27. sota / sote- *sota ‘war’ *sota-i- ‘wars’ 28. pora / pore- *pora ‘drill’ *pora-i- ‘drills’ 29. muna /mune- *muna ‘egg’ *muna-i- ‘eggs’ 30. rulla /rulle- *rulla ‘roll’ *rulla-i- ‘rolls’ 31. tupa / tupe- *tupa ‘cabin’ *tupa-i- ‘cabins’ 32. jyvä / jyve- *juvä ‘grain’ *jyvä-i- ‘grains’ 33. hätä /häte- *hätä ‘distress’ *häta-i- ‘distresses’ 34. mökä / möke- *mökä ‘hullabaloo’ *möka-i- ‘hullabaloos’ Changes: Change 1.: V > Ø /V __ +i Complex vowel nuclei were adjusted so that no more than two moras appear; thus, long vowels were shortened and the last vowel of diphthongs (vowel clusters) was dropped before i, presumably to prevent more than two moras from coming together), in examples 1.-25., for example. Change 2. a-raising and dissimilation: a > o /a, e, i Co __ +i; a > e /o, u Co __ +i; ä > e /y, ö, ä Co __ +i. That is, before i in a suffix, a was raised to a mid vowel, either to e or o, the choice of which was determined by dissimilation from the vowel of the preceding syllable. Thus if the vowel of the preceding syllable is back and round (as in 27. sota ‘war’), the second vowel raises and dissimilates to be non-back and non-round, *sota-i- > sotei (and eventually on to soti- in modern Finnish), and if the preceding vowel is non-round (a, e, i), then the a in this context raises and dissimilates to a back round vowel (as in 22. where *sata-i- ‘hundreds’ > sato-i-). Where ä is the vowel that changes, it raises to e before -i). More formally, the change might be stated as something like: a > [-low, -αback, -βround] /[αback, βround]VCo __+i ä > [-low] /VCo __ +i Change 3. Long-Mid Vowel Diphthongization: ee > ie, oo > uo, öö > yö.
Examples of the reconstructions and how the changes apply to them: 9. *soo *soo-i 27. *sota *sota-i 22. *sata *sata-i- ‘grants’ ‘granted’ ‘war’ ‘wars’ ‘hundred’ ‘hundreds’ Change 1. -- soi ------Change 2. ------sotei -- satoi Change 3. suo ------Result: suo soi sota sotei- sata satoi-
Relative Chronology: As stated, Change 1., which works to prohibit vowel clusters of more than two moras, must have taken place before Change 3., diphthongization. Otherwise, we would get erroneous results, if the changes had happened in the opposite order:
99 9. *soo *soo-i ‘grants’ ‘granted’ Change 3. suo suo-i Change 1. -- sui Result: suo ✘sui
Exercise 8.5 Nahuatl internal reconstruction
Nahuatl is a Uto-Aztecan language, spoken by over 1,000,000 people in Mexico; it was the language of the Aztecs and the Toltecs. Compare the following words. Find the forms which have variants; apply internal reconstruction to these forms. Reconstruct a single original form for the morphemes which have alternate shapes, and postulate the changes which you think must have taken place to produce these variants. Can you establish a relative chronology for any of these changes? Present your reasoning; why did you choose this solution and reject other possible hypotheses? NOTE: tl is a single consonant, a voiceless lateral affricate; k w is a labialized velar stop and is a single segment; č = IPA tʃ; š = IPA ʃ. Note that the morpheme which has the allomorphs -tl, -tli, -li is traditionally called the ‘absolutive’; it has no other function than to indicate a noun root which has no other prefixes or suffixes.
1a tepos-tli ‘axe’ 13a ikʃi-tl ‘foot’ 1b no-tepos ‘my axe’ 13b no-kʃi ‘my foot’ 1c tepos-tlān ‘place of axes’ 14a ikni-tl ‘fellow’ 2a kak-tli ‘shoe, sandal’ 14b no-kni ‘my fellow’ 2b no-kak ‘my shoe, sandal’ 15a isti-tl ‘fingernail’ 3a teʃ-tli ‘flour’ 15b no-sti ‘my fingernail’ 3b no-teʃ ‘my flour’ 16a ihti-tl ‘stomach’ 4a mis-tli ‘cougar’ 16b n-ihti ‘my stomach’ 4b mis-tlān ‘place of cougars’ 17a īʃte-tl ‘eye’ 5a kal-li ‘house’ 17b n-īʃte ‘my eye’ 5b no-kal ‘my house’ 18a ihwi-tl ‘feather’ 6a tlāl-li ‘land’ 18b n-ihwi ‘my feather’ 6b no-tlāl ‘my land’ 19a itskwin-tli ‘little dog’ 7a čīmal-li ‘tortilla griddle’ 19b n-itskwin ‘my little dog’ 7b no-čīmal ‘my tortilla griddle’ 20a ička-tl ‘cotton’ 7c čīmal-lan ‘place of 20b no-čka ‘my cotton’ tortilla griddles’ 8a mīl-li ‘cornfield’ 21a okič-tli ‘male, man’ 8b no-mīl ‘my cornfield’ 21b n-okič ‘my husband’ 8c mīl-lan ‘place of cornfields’ 22a kaʃi-tl ‘bowl’ 9a āma-tl ‘paper, fig tree’ 22b no-kaʃ ‘my bowl’ 9b n-āma ‘my paper, fig tree’ 23a kwawi-tl ‘tree, wood’ 9c āma-tlān ‘place of paper, 23b no-kwaw ‘my tree, wood’ fig trees’ 10a e-tl ‘bean’ 24a māyi-tl ‘hand’ 10b n-e ‘my bean’ 24b no-māy ‘my hand’
100 10c e-tlān ‘place of beans’ 25a ʃāmi-tl ‘brick’ 11a siwā-tl ‘woman’ 25b no-ʃān ‘my brick’ 11b no-siwā ‘my wife’ 26a pāmi-tl ‘flag’ 11c siwa-tlān ‘place of women’ 26b no-pān ‘my flag’ 12a ol-li ‘rubber’ 27a kōmi-tl ‘jug’ 12b n-ol ‘my rubber’ 27b no-kōn ‘my jug’ 12c ol-lān ‘place of rubber’
Reconstructions of Pre-Nahuatl: Affixes: *-tlān ‘place of’ *-tli ‘Absolutive’ *no- ‘my’ Roots: 1 *tepos ‘axe’ 15 *sti ‘fingernail’ 2 *kak ‘shoe, sandal’ 16 *ihti ‘stomach’ 3 *teš ‘flour’ 17 *īšte ‘eye’ 4 *mis ‘cougar’ 18 *ihwi ‘feather’ 5 *kal ‘house’ 19 *itskwin ‘little dog’ 6 *tlāl ‘land’ 20 *čka ‘cotton’ 7 *čīmal ‘tortilla griddle’ 21 *okič ‘male, man’ 8 *mīl ‘cornfield’ 22 *kaši ‘bowl’ 9 *āma ‘paper, fig tree’ 23 *kwawi ‘tree, wood’ 10 *e ‘bean’ 24 *māyi ‘hand’ 11 *siwā ‘woman’ 25 *šāmi ‘brick’ 12 *ol ‘rubber’ 26 *pāmi ‘flag’ 13 *kši ‘foot’ 27 *kōmi ‘jug’ 14 *kni ‘fellow’
Changes: Change 1. tl-deaffrication: tl > l /l __ The lateral affricate tl became l after another l, i.e. tl-l > ll (as in 5a, 6a, 7a, 7c, 8a, 8c, 12a, 12c) Change 2. o-deletion: o > Ø /n__+V (as in 9b, 10b, 12b, 16b, 17b, 18b, 19b) The o of prefixes of the shape no- was deleted when the prefix was attached to a stem beginning in a vowel. Change 3. Epenthesis: Ø > i /#__CC (as in 13a, 14a, 15a, 20a) Word-initially before a consonant cluster, i was inserted. Change 4. Apocope of final i: i > Ø /VC__# (as in 22b, 23b, 24b, 25b, 26b, 27b) A final i was deleted in polysyllabic words when the i was preceded by a single consonant. Change 5. Final m > n:
101 m > n /__ # (as in 25b, 26b, 27b). Final m became n.
Examples of reconstructions with changes applied to them in temporal order: 1a *tepos-tli 5a *kal-tli 7c *čīmal-tlān 9b *no-āma ‘axe’ ‘house’ ‘tortilla griddle’ ‘my paper’ Change 1. -- kal-li čīmal-lān -- Change 2. ------n-āma Change 3. ------Change 4. ------Change 5. ------Result: tepostli kalli čīmallān nāma
13a *kši-tli 15b *no-kši 16a *ihti-tli 16b *no-ihti ‘foot’ ‘my foot’ ‘stomach’ ‘my stomach’ Change 1. ------Change 2. ------n-ihti Change 3. ikši-tli ------Change 4. ikši-tl -- ihti-tl -- Change 5. ------Result: ikšitl nokši ihtitl nihti
22a *kaši-tli 22b *no-kaši 25a *šāmi-tli 25b *no-šāmi ‘bowl ‘my bowl’ ‘brick’ ‘my brick’ Change 1. ------Change 2. ------Change 3. ------Change 4. kaši-tl no-kaš šāmi-tl no-šām Change 5. ------no-šān Result: kašitl nokaš šāmitl nošān
Relative chronology: Change 1. tl-deaffrication does not interact with the other changes and therefore could have taken place at any time relative to any of the other changes. Change 4. Apocope of final i took place before Change 5. (final m > n), for otherwise we would get erroneous results if we assumed they took place in the opposite temporal sequence, as in, for example: 25a *šāmi-tli 25b *no-šāmi ‘brick’ ‘my brick’ Change 5. -- -- Change 4. šāmi-tl no-šām Result: šāmitl ✘nošām As stated, Change 2. o-deletion and Change 3. i-epenthesis do not interact with one another; however, if formulated differently, one might get erroneous results if the changes had not taken place in the order of 2. before 3. If Change 3 did not specify word-initially before a consonant cluster (/# __ CC) but were formulated, for example, as though it applied before a
102 consonant cluster at the beginning of a morpheme (/+__CC), then the temporal order would need to be stipulated, as seen for example in the following where the changes are in a different sequence and produce erroneous results: 13a *kši-tli 15b *no-kši 16a *ihti-tli 16b *no-ihti ‘foot’ ‘my foot’ ‘stomach’ ‘my stomach’ Change 3. ikši-tli no-ikši -- -- Change 2. -- n-ikši -- n-ihti Change 4. ikši-tl -- ihti-tl -- Result: ikši-tl ✘nikši ihtitl nihti
Exercise 8.6 Indonesian internal reconstruction
Identify the morphemes which have more than one variant in the following data from Indonesian (an Austronesian language). Apply internal reconstruction to these forms; reconstruct a single original form for each of the roots and for the prefix, and postulate the changes you think must have taken place to produce these variants. Can you establish a relative chronology for any of these changes? Provide sample derivations which show your reconstruction and how the changes apply to it for both the simple and the prefixed forms in 2, 12, 13, 15 and 19. (The prefix in the second column has a range of functions, among them, it places focus on the agent (‘doer’) of a verb, derives transitive or causative verbs, and derives verbs from nouns.) HINT: relative chronology is important to the solution of this problem. (/ɲ/ = palatal nasal.)
simple form prefixed form gloss 1. lempar məlempar ‘throw’ 2. rasa mərasa ‘feel’ 3. wakil məwakil- ‘represent’ 4. yakin məyakin- ‘convince’ 5. masak məmasak ‘cook’ 6. nikah mənikah ‘marry’ 7. ŋačo məŋačo ‘chat’ 8. ɲaɲi məɲaɲi ‘sing’ 9. hituŋ məŋhituŋ ‘count’ 10. gambar məŋgambar ‘draw a picture’ 11. kirim məŋirim ‘send’ 12. dəŋar məndəŋar ‘hear’ 13. tulis mənulis ‘write’ 14. bantu məmbantu ‘help’ 15. pukul məmukul ‘hit’ 16. ǰahit məɲǰahit ‘sew’ 17. čatat məɲčatat ‘note down’ 18. ambil məŋambil ‘take’ 19. isi məŋisi ‘fill up’ 20. undaŋ məŋundaŋ ‘invite’
103 Identify the morphemes which have more than one variant: 11. kirim / -irim ‘send’ (in 11.) 13. tulis / -ulis ‘hear’ (in 13.) 15. pukul / -ukul ‘hit’ (in 15.) Prefix: mə- (in 1.-8.), məŋ- (in 9.-11., 18.-20.), mən- (in 12.-13.), məm- (in 14.-15.), məɲ- (in 16.-17.)
Reconstruct a single original form for each of the roots and for the prefix: 11. *kirim ‘send’ 13. *tulis ‘hear’ 15. *pukul ‘hit’ Prefix *məŋ-
Postulate the changes: Change 1. Nasal loss: Nasal > Ø /__ Sonorant (as in 1.-8.) Nasals were deleted before sonorants (l, r, w, y, other nasals). Change 2. Nasal assimilation: Nasal > [αPlace] /__ [αPlace, non-continuent] Nasals assimilate to the point of articulation of a following stop or affricate: ŋ > m /__ p, b (in 14.-15.) ŋ > n /__ t, d (in 12.-13.) ŋ > ɲ /__ č, ǰ (in 16.-17.) ŋ > ŋ /__ k, g (here the change applies only vacuously; since it starts out as ŋ already, assimilating it to the place of articulation of velar stops leaves it as ŋ) (in. 10.-11.) Change 3. Voiceless stop loss after a nasal: Voicless stop > Ø /Nasal __ Voiceless stops were lost after nasals (as in 11., 13., 15.)
Provide sample derivations which show your reconstruction and how the changes apply to it for both the simple and the prefixed forms in 2, 12, 13, 15 and 19. 2. ‘feel’ 12. ‘hear’ *rasa *məŋ-rasa *dəŋar *məŋ-dəŋar Change 1. Nasal-loss: -- mə-rasa -- -- Change. 2. Nasal assimilation: ------mən-dəŋar Change 3. Voiceless stop loss: ------Result: rasa mə-rasa dəŋar məndəŋar
13. ‘write’ 15. ‘hit’ *tulis *məŋ-tulis *pukul *məŋ-pukul Change 1. Nasal-loss: ------Change. 2. Nasal assimilation: -- mən-tulis -- məm-pukul Change 3. Voiceless stop loss: -- mənulis -- məm-ukul Result: tulis mənulis pukul məmukul
104 19. ‘fill up’ 20. ‘invite’ *isi *məŋ-isi *undaŋ *məŋ-undaŋ Change 1. Nasal-loss: ------Change. 2. Nasal assimilation: ------Change 3. Voiceless stop loss: ------Result: isi məŋisi undaŋ məŋundaŋ
Relative chronology: Change 1. Nasal loss does not interact with the other changes and so could have taken place at any time with respect to the other two changes. However, Change 2. (Nasal assimilation) had to have taken place before Change 3. (Voiceless stop loss), for otherwise erroneous forms would result, as seen, for example, in the following, if the temporal sequence were assumed to be in the other order, with 3. before 2.: 13. ‘write’ 15. ‘hit’ *tulis *məŋ-tulis *pukul *məŋ-pukul Change 3. Voiceless stop loss: -- məŋ-ulis -- məŋ-ukul Change. 2. Nasal assimilation: ------Result: tulis ✘məŋulis pukul ✘məŋukul
Exercise 8.7 Tol (Jicaque) internal reconstruction
Jicaque (called Tol by its speakers) is spoken in Honduras. State the variants (allomorphs) of the roots and of the possessive pronominal prefixes; apply internal reconstruction to these forms. Reconstruct a single original form for each root morpheme and write the changes which you think must have taken place to produce these variants. Present your reasoning; why did you choose this solution and reject other possible hypotheses? (HINT: the original form of the possessive pronouns was: *n- ‘my’, *hi- ‘your’, *hu- ‘his’; original *n+h > n.) Note that what is structurally a labialized w is realized phonetically as [w_], but is written as ww in this problem.
my your his Meaning of the noun root 1. mbata peta pota ‘duck’ 2. mbapay pepay popay ‘father’ 3. ndaʔ teʔ toʔ ‘man’s brother’ 4. ndarap terap torap ‘woman’s younger sister’ 5. ŋkhan khen khon ‘bed’ 6. nlara lera lora ‘mouth’ 7. ntsham tshem tshom ‘foot’ 8. mbe hepe pwe ‘rock, stone’ 9. mbep hepep pwep ‘fingernail’ 10. mberam heperam pweram ‘tongue’ 11. mphel hephel phwel ‘arm’ 12. ŋgerew hekerew kwerew ‘cousin’ 13. ŋkhere hekhere khwere ‘bone’ 14. ŋgiway hikiway kwiway ‘woman’s brother’ 15. nǰič hičič čwič ‘tendon’
105 my your his Meaning of the noun root 16. nǰipe hičipe čwipe ‘paired sibling’ 17. mbomam pyomam hopomam ‘chokecherry’ 18. mphok phyok hophok ‘cheek’ 19. ŋgol kol hokol ‘belly’ 20. nts’ul ts’yul huts’ul ‘intestines’ 21. mphɨya phyeya hɨphɨya ‘tobacco’ 22. mp’ɨs p’yɨes hɨp’ɨs ‘deer’ 23. ndɨm tyem hɨtɨm ‘heel’ 24. mbasas wesas wosas ‘woman’s sister-in-law’ 25. mbis hiwis wwis [wɨis] ‘tooth’ 26. mbin hiwin wwin [wɨin] ‘toad’ 27. mboyum wyoyum howoyum ‘husband’ 28. namas mes mos ‘hand’ 29. nemen hemen mwen ‘neck’ 30. nimik himik mwik ‘nose’ 31. nɨmɨnɨ mjenɨ hɨmɨnɨ ‘yam’ 32. namap hemap homap ‘aunt’ 33. nasunu hesunu hosunu ‘chest’
Variants (allomorphs) of the roots (together with changes in roots that reflect possessive pronouns): 1. -bata / peta / pota 2. -bapay / pepay / popay 3. -daʔ / teʔ / toʔ 4. -darap/ terap / torap 5. -khan / khen / khon 6. -lara / lera / lora 7. -tsham / tshem / tshom 8. -be / -pe / pwe 9. -bep / -pep / pwep 10. -beram / -peram / pweram 11. -phel / phwel 12. -gerew / -kerew / kwerew 13. -khere / khwere 14. -giway / -kiway kwiway 15. -ǰič / / -čič 16. -ǰipe / -čipe / čwipe 17. -bomam / pyomam / -pomam 18. phok / phyok 19. -gol / kyol / -kol 20. -ts’ul / ts’yul 21. phɨya / phyeya 22. -p’ɨs / p’yes 23. dɨm / tyɨm / tɨm 24. -basas / wesas / wosas 25. -bis / -wis / -wwis 26. -bin / -win / wwin 27. -boyum / wyoyum / -woyum
106 28. -mas / -mes / -mos 29. -men / mwen 30. -mik / mwik 31. -mɨnɨ / myenɨ 32. -amap / hemap / homap 33. -asunu / hesunu / hosunu
Variants of the possessive pronominal prefixes: ‘my’: m- /__ labials, n- /__ alveolars, ŋ /__ velars (that is, a nasal which assimilates to the point of articulation of the following consonant), nV1 /__+ Nasal V1 (the prefix is nV-, where the vowel is identical to the first vowel of the root, when prefixed to a nasal-initial root, as in 28., 29., 30., 31.). ‘your’: palatalization of the initial consonant of forms which have o, u, or ɨ as the first vowel of the root (as in 17., 20., 21.-23., 27., 31.), change of a or ɨ when they are the first vowel of a root to e (as in 1.-7., 21.,-24., 28., 31.-33.), hi- when the first vowel of the root to which this prefix is attached is i, and he- when the first vowel of the root is e (as in 8.,-16., 25.-26., 29.). ‘his’: labialization of the initial consonant of forms which have e or i as the first vowel of the root (as in 8.-16., 25.-26., 29.-30.), change of a to o when a is the first vowel of a root (as in 1.-7., 24., 28., 32.-33.), ho- with o as the first vowel of the root to which this prefix is attached, hu with u, and hɨ with ɨ (as in 17.-23., 27., 31.)
Reconstructions: *n- ‘my’ *hi- ‘your’ *hu- ‘his’ 1. *pata ‘duck’ 2. *papay ‘father’ 3. *taʔ ‘man’s brother’ 4. *tarap ‘woman’s younger sister’ 5. *khan ‘bed’ 6. *lara ‘mouth’ 7. *tsham ‘foot’ 8. *pe ‘rock, stone’ 9. *pep ‘fingernail’ 10. *peram ‘tongue’ 11. *phel ‘arm’ 12. *kerew ‘cousin’ 13. *khere ‘bone’ 14. *kiway ‘woman’s brother’ 15. *čič ‘tendon’ 16. *čipe ‘paired sibling’ 17. *pomam ‘chokecherry’ 18. *phok ‘cheek’ 19. *kol ‘belly’ 20. *ts’ul ‘intestines’ 21. *phɨya ‘tobacco’ 22. *p’ɨs ‘deer’ 23. *tɨm ‘heel’ 24. *wasas ‘woman’s sister-in-law’ 25. *wis ‘tooth’
107 26. *win ‘toad’ 27. *woyum ‘husband’ 28. *mas ‘hand’ 29. *men ‘neck’ 30. *mik ‘nose’ 31. *mɨnɨ ‘yam’ 32. *hamap ‘aunt’ 33. *hasunu ‘chest’
Changes: Change 1 Nasal-assimilation: n > [αPlace] /__ [αPlace, -nasal, -low]C Nasals assimilate to point of articulation of a following consonant when that C is not nasal or h. As in 1. *n-pata ‘my duck’ > m-pata (later mbata), 3. *n-taʔ ‘man’s brother’ > ntaʔ (later *ndaʔ), 5. *n-khan > ŋkhan ‘bed’, also in 2., 4., 6.-27. Change 2 Voicing: p, t, k, č > voiced /Nasal __ Plain stops and affricates (non-aspirated and non-glottalized ones) are voiced after a nasal, as in 1. *n-pata ‘my duck’ > mpata (and to mbata later). (See also 2.-4., 8.10., 12., 14.-17., 19., 23.) Change 3 h-loss: h > Ø /n __ As in 32. *n-hamap ‘aunt’ > namap, also in 33. Change 4 w > b: w > b /Nasal__. A w changes to b when following a nasal, as in 25. *n-wis ‘my tooth’ > m-wis by Change 1 (nasal-assimilation) > mbis by Change 4. (See also 24.-27.) Change 5a Palatalization: C > Cy /i+ __ V (V ≠ i). A consonant is palatalized after i of a prefix if the following vowel is not i, as in 1. *hi-pata ‘your duck’ > hi-pyata (and by later changes > peta), 20. *hi-ts’ul ‘your intestines’ > hits’yul (and by later changes to ts’yul), and in 2.-7., 17.-19., 21.-24., 27.-28., 31. Change 5b Labialization: C > Cw /u + __ V (V ≠ u, o). A consonant is labialized after u of a prefix if the following vowel is not round (not u or o) as in 1. *hu-pata ‘his duck’ > hu-pwata (and by later changes pota), 8. *hu-pe ‘his rock, stone’ > hu-pwe (and by later changes pwe), and in 2.-16., 24.-26., 28.-30. Note that 5a and 5b can be seen as a single change (collapsed together by the notation of distinctive features, though the classes for the vowels conditioning labialization and palatalization are difficult to reconcile). Change 6a Vowel assimilation of i: i > e /h__ +Ce As in 8. *hi-pe ‘your rock, stone’ > hepe; see also 9.-13., 29. Change 6b Vowel assimilation of u: u > o /h__ +Co, u > ɨ /h__ +Cɨ As in 19. *hu-kol ‘his belly’ > hokol, and in 22. *hu-p’ɨs ‘his deer’ > hɨp’ɨs; see also 17., 21., 22.-23., 27., 31. That is, the i of the prefix hi- assimilates to e when the first vowel of the root to which this prefix is attached has e, and the u of hu- assimilates to o when the first vowel of the root to which the prefix is joined has o, and to ɨ when the first vowel of the root to which the prefix is joined has ɨ. Note that changes 6a and 6b can be combined into a single change using the distinctive features of the vowels involved. Change 7 hi- and hu- loss: hi- and hu- > Ø /__ Cy, Cw. After the first consonant of the root was palatalized (in ‘your’ forms due to the influence of the vowel in the prefix hi-) or labialized (in ‘his’ forms due to the influence of the vowel in
108 the prefix hu-), the palatalization and the labialization were sufficient to mark these forms as ‘your’ or ‘his’ respectively, and so the now redundant prefixes hi- and hu- were dropped. This is seen for example in: 1. *hi-pata ‘your duck’ > hipyata by Change 5a (palatalization) > pyata by Change 7 (and peta by later changes) and *hu-pata ‘his duck’ > hupwata by Change 5b (labialization) > pwata by Change 7 (and pota by later changes). In 8. *hu-pe ‘his rock, stone’ > hupwe by Change 5b, then pwe by Change 7, but *hi-pe ‘your rock, stone’ does not fit the environment for Change 5a (palatalization), and so the prefix *hi- remains, though the vowel of the prefix is assimilated, turning it into hepe by Change 6a. (See also 2.-7., 9.-16., 24.-26., 28.-30.) It is unusual to delete multiple segments at once in a single sound change; it is possible that this involved intermediate stages, perhaps first loss of the vowel i or u of the prefixes when the following consonant was palatalized or labialized and then later loss of the h, or the reverse, first the lost of h followed later by the loss of i- and u- (*hu-pe ‘his stone’ > hupwe > hpwe > pwe ?). Change 8 Vowel coalescence: Cya, Cyɨ > Ce; Cwa > Co. The palatal off-glide that results from Change 5a and an a or ɨ next to it together coalesce (like monophthongization) into e, and the labial off-glide that results from Change 5b and a following a coalesce to o. This is seen in, for example 1. hi-pata ‘your duck’ > hipyata by Change 5a, then hipyata > pyata by Change 6, and finally pyata > peta by rule 8; similarly 1. hu-pata ‘his duck’ > hupwata by Change 5b, then hupwata > pwata by Change 7, and finally pwata > pota by rule 8. (See also 2.-7., 21.-24., 28., 31.-33.) Change 9: ww > wɨ This is a phonetic change to accommodate the structure of labialization for ‘his’ forms where there is the phonetic problem that it is not actually possible to labialize a w, which is already labial, seen in 25. *hu-wis ‘his tooth’ > hu-wwis (by Change 5b.) > wwis by Change 7. and this wwis > [wɨis] by Change 9. (See also 26.) Change 10 Epenthesis: Ø > V1 /n__ Nasal V1. A vowel which is a copy of the first root vowel is inserted between n- of a prefix and a nasal, as in 28. *n-mas ‘my hand’ > namas. (See also 29.-30.)
Derivations: the following examples illustrate how these changes applied to various reconstructed forms. 1. ‘my duck’ ‘your duck’ ‘his duck’ *n-pata *hi-pata *hu-pata Change 1 mpata -- -- Change 2 mbata -- -- Change 3 ------Change 4 ------Change 5 -- hipyata hupwata Change 6 ------Change 7 -- pyata pwata Change 8 -- peta pota Change 9 ------Change 10 ------Result: mbata peta pota
109
2. ‘my father’ ‘your father’ ‘his father’ *n-taʔ *hi-taʔ *hu-taʔ Change 1 ntaʔ -- -- Change 2 ndaʔ -- -- Change 3 ------Change 4 ------Change 5 -- hityaʔ hutwaʔ Change 6 ------Change 7 -- tyaʔ twaʔ Change 8 -- teʔ toʔ Change 9 ------Change 10 ------Result: ndaʔ teʔ toʔ
13. ‘my bone’ ‘your bone’ ‘his bone’ *n-khere *hi-khere *hu-khere Change 1 ŋkhere -- -- Change 2 ------Change 3 ------Change 4 ------Change 5 -- -- hukhwere Change 6 -- hekhere -- Change 7 -- -- khwere Change 8 ------Change 9 ------Change 10 ------Result: ŋkhere hekhere khwere
15. ‘my tendon’ ‘your tendon’ ‘his tendon’ *n-čič *hi-čič *hu-čič Change 1 (nčič) -- -- Change 2 nǰič -- -- Change 3 ------Change 4 ------Change 5 -- -- hučwič Change 6 ------Change 7 -- -- čwič Change 8 ------Change 9 ------Change 10 ------Result: nǰič hičič čwič
110 19. ‘my belly’ ‘your belly’ ‘his belly’ *n-kol *hi-kol *hu-kol Change 1 ŋkol -- -- Change 2 ŋgol -- -- Change 3 ------Change 4 ------Change 5 -- hikyol -- Change 6 -- -- hokol Change 7 -- kyol -- Change 8 ------Change 9 ------Change 10 ------Result: ŋgol kyol hokol
25. ‘my tooth’ ‘your tooth’ ‘his tooth’ *n-wis *hi-wis *hu-wis Change 1 mwis -- -- Change 2 ------Change 3 ------Change 4 mbis -- -- Change 5 -- -- huwwis Change 6 ------Change 7 -- -- wwis Change 8 ------Change 9 -- -- wɨis Change 10 ------Result: mbis hiwis wɨis
29. ‘my neck’ ‘your neck’ ‘his neck’ *n-men *hi-men *hu-men Change 1 ------Change 2 ------Change 3 ------Change 4 ------Change 5 -- -- humwen Change 6 -- hemen -- Change 7 -- -- mwen Change 8 ------Change 9 ------Change 10 nemen -- -- Result: nemen hemen mwen
111 33. ‘my chest’ ‘your chest’ ‘his chest’ *n-hasunu *hi-hasunu *hu-hasunu Change 1 ------Change 2 ------Change 3 nasunu -- -- Change 4 ------Change 5 -- hihyasunu huhwasunu Change 6 ------Change 7 -- hyasunu hwasunu Change 8 -- hesunu hosunu Change 9 ------Change 10 nemen -- -- Result: nasunu hesunu hosunu
Exercise 8.8 Samoan internal reconstruction
Compare words in the two columns. Identify the morphemes which have more than one variant. Reconstruct a single original form for all the morphemes here, and postulate the changes you think must have taken place to produce these forms. Can you establish a relative chronology for any of these changes? Why did you choose this solution and reject other possible hypotheses? NOTE: many roots will have two allomorphs; the suffix also has more than one variant in several of the cases. HINT: think ‘predictability’ and exploit gaps in the inventory of the consonants that can occur as the last consonant in these words.
1. alofa ‘love’ alofaŋia ‘loved’ 2. taŋo ‘grasp’ taŋofia ‘grasped’ 3. fua ‘measure’ fuatia ‘measure’ 4. au ‘reach’ aulia ‘reached’ 5. faitau ‘read’ faitaulia ‘read’ 6. u ‘to bite’ utia ‘bitten’ 7. ula ‘smoke’ ulafia ‘smoked’ 8. na ‘hide’ natia ‘hidden’ 9. fau ‘bind’ fausia ‘bound’ 10. ʔata ‘laugh’ ʔataŋia ‘laughed’ 11. inu ‘drink’ inumia ‘drunk’ 12. taofi ‘hold’ taofia ‘held’ 13. mu ‘burn’ muina ‘burned’ 14. tuʔu ‘put’ tuʔuina ‘put’ 15. faŋa ‘feed’ faŋaina ‘fed’ 16. sauni ‘prepare’ saunia ‘prepared’ 17. siʔi ‘raise’ siʔitia ‘raised’ 18. pisi ‘splash’ pisia ‘splashed’ 19. ao ‘gather’ aofia ‘gathered’ 20. ilo ‘perceive’ iloa ‘perceived’
112 21. ʔave ‘take’ ʔavea ‘taken’ 22. oso ‘jump’ osofia ‘jumped’ 23. ʔino ‘hate’ ʔinosia ‘hated’ 24. filo ‘mix’ filoŋia ‘mixed’ 25. fasioti ‘kill’ fasiotia ‘killed’ 26. utu ‘fill’ utufia ‘filled’ 27. ufi ‘cover’ ufitia ‘covered’ 28. ʔai ‘eat’ ʔaiina ‘eaten’ (cf. 6) 29. afio ‘come in’ afioina ‘(has) come in’ 30. laʔa ‘step over’ laʔasia ‘stepped over’ 31. manaʔo ‘want’ manaʔomia ‘wanted’ 32. mataʔu ‘destroy’ mataʔutia ‘destroyed’ 33. milo ‘twist’ milosia ‘twisted’ 34. taŋi ‘cry’ taŋisia ‘cried’ 35. vavae ‘divide’ vavaeina ‘divided’
This is an unusual problem. The solution is relatively simple once found, but it may be difficult to discover. To solve it, first attempt to determine the allomorphs of the roots and suffixes. A clue to the morphological analysis is: seek “recurrent partials” – recurring parts of the words that seem to have the same meaning of function. In looking for recurrent partials, you might proceed in two different ways for the data in this exercise. First, if you compare across the columns, you start seeing what in the first column is repeated as a recurrent partial in the second column, with something left over in the second column. For example, in the first row alofa occurs in both columns, with -ŋia left over in the second column. In this sort of comparison across columns, we see: alofa -ŋia taŋo -fia fua -tia au -lia etc. This kind of comparison between the forms in each row would seem to find that the root in both columns is constant and would postulate what is left over as a suffix, but a suffix with a large number of alternants, -ŋia (in 1., 10., 24.), -fia (in 2., 7., 12., 19., 22.), -tia (in 3., 6., 8., 17., 27., 32.), -lia (in 4., 5.), -sia (9., 23., 30., 33., 34.), -mia (11., 31.), etc. Since nothing about the final V of the root nor about the consonant before it or anything else in the phonology of the roots seems to condition differences in the suffixes, to sustain this hypothesis of multiple allomorphs for this suffix, it would be necessary to assume that Samoan children acquire the language by just memorizing which from the various suffix forms goes with each root, alofa taking -ŋia, taŋo requiring -fia, and so on. That line of reasoning would postulate morphological conditioning, as in English where speakers memorize that the plural allomorph is -en with ox, and Ø with deer, rather than the more usual phonologically conditioned allomorphs of plural, where the form is predictable based on the phonology of the forms to which they are attached, with ɨz as the allomorph of plural after sibilants as in /fɔksɨz/ ‘foxes’, voiceless s after voiceless consonants as in /kæt-s/ ‘cats’, and voiced after voiced consonants as in /dɔgz/ ‘dogs’. Morphological conditioning is unusual. This solution would require a very heavy learning burden for Samoan children who would have to memorize which different form of the suffix
113 goes with each root. In the second kind of comparison, if you compare the forms in the second column with one another, what recurs in most cases is -ia, which leaves seemingly different forms of the roots in the first and second columns, the form in the first column with no final consonant, and the form in the second often with a final consonant: 1. alofa lofaŋ- + -ia 2. taŋo taŋof- + -ia 3. fua fuat- + -ia etc. In this approach to the data, we would reconstruct these roots with the final consonant present together with a postulated sound change that deletes final consonants when no suffix is present to prevent them from occurring in final position.
Sound change (1) Final consonant loss: C > Ø /__ # As in: Pre-Samoan *alofaŋ ‘love’ *alofaŋ-ia ‘loved *taŋof ‘grasp *taŋof-ia ‘grasped’ Change (1): alofa -- taŋo -- Result: alofa alofaŋia taŋo taŋofia Not all roots appear to have had final consonants, however. Some appear to have no final consonant and some also seem to take a different form of the suffix, as in: 16. sauni saunia 18. pisi pisia 20. ilo iloa 21. ʔave ʔavea 25. fasioti fasiotia In these cases, the sufix *-ia is added to a vowel-final root and the i of the suffix is deleted (presumably to avoid vowel clusters of three members); we postulate: Sound change (2) i-deletion in vowel clusters: i > Ø /V+__a This change is seen in the following:
Pre-Samoan *sauni ‘prepare’ *sauni-ia ‘prepared’ *ilo ‘perceive’ *ilo-ia ‘perceived’ Change (2): -- saunia -- iloa Result: sauni saunia ilo iloa However, not all forms appear to behave as those involving consonant-final roots subject to change (1) or vowel-final roots subject to change (2). A number of seeming exceptions remain: 13. mu muina 14. tuʔu tuʔuina 15. faŋa faŋaina 28. ʔai ʔaiina 29. afio afioina 35. vavae vavaeina These forms present a problem for the analysis so far, for if we assume they are vowel-final, we would expect mu / mua and afio / afioa, or we would expect mu and afio to have a consonant when followed by the suffix, muC-ia and afioC-ia (where C is some unpredictable consonant that is part of the root). A look at the final consonant of the roots which do show up before -ia in
114 the data reveals C+ia combinations with all possible consonants of Samoan except -n+ia. We take advantage of the gap in the inventory of root-final consonants that occur, from which -n is missing, together with the problem of otherwise not being able to predict the outcomes for these exceptional forms, and we postulate an original root-final -n in these forms and a sound change of metathesis: Change (3) n-metathesis: n+ia > ina Seen in the following: Pre-Samoan *mun ‘burn *mun-ia ‘burned *afion ‘come in’ *afion-ia ‘entered’ Change (1): mu -- afio -- Change (2) ------Change (3) -- muina -- afioina Result: mu muina afio afioina The internal reconstructions postulated for the different sorts of roots and and how the sound changes apply to them are illustrated in the following forms: Pre-Samoan *alofaŋ *alofaŋ-ia *ilo *ilo-ia *mun *mun-ia ‘love’ ‘loved ‘perceive’ ‘perceived’ ‘burn’ ‘burned Change (1): alofa ------mu -- Change (2) ------iloa -- -- Change (3) ------muina Result: alofa alofaŋia ilo iloa mu muina
Exercise 8.9 Nivaclé internal reconstruction
Nivaclé (also called Chulupí) is a Matacoan language of northern Argentina and Paraguay. Compare the forms in the two columns; identify the morphemes which have more than one phonological shape (variant). Attempt to reconstruct a unique form for each of these words in Pre-Nivaclé, and state the changes which have taken place, according to your analysis, in the transition from Pre-Nivaclé to modern Nivaclé. Is there any relative chronology involved in the changes you postulate? If so, state what it is and show sample derivations of at least four word pairs (for example both words of 12, 14, 18, and 20). NOTE: /kl/ is a single segment, both phonemically and phonetically – the velar closure and the lateral articulation are released simultaneously as a single sound; /ts/ is also a single segment, an alveolar affricate; /ɑ/ is a back low vowel, and contrasts with /a/, a cental low vowel; ɫ = voiceless l; C’ = glottalized [ejective] consonants. The š is IPA ʃ, the č IPA tʃ. The plural suffixes, in the second column, have several different forms, but do not attempt to reconstruct them.
Set I 1. xutsax ‘vulture’ xutsx-as ‘vultures’ 2. ɸatsux ‘centipede’ ɸatsx-us ‘centipedes’ 3. snomax ‘ash’ snomx-as ‘ashes’ 4. ɫtɑsex ‘seed’ ɫtɑsx-ey ‘seeds’ 5. kutsxanax ‘thief’ kutsxanx-as ‘thieves’ 6. ipɑset ‘my lip’ ipɑst-es ‘my lips’ 7. nasuk ‘guayacán (tree)’ nask-uy ‘guayacans’
115 8. ɸaʔayuk ‘algarrobo (tree)’ ɸaʔayk-uy ‘algarrobos’ 9. axɑyuk ‘mistol (tree)’ axɑyk-uy ‘mistols’
Set II 10. inkɑʔp ‘year’ inkɑp-es ‘years’ 11. ɫuʔp ‘nest’ ɫup-is ‘nests’ 12. k’utxaʔn ‘thorn’ k’utxan-is ‘thorns’ 13. ɫsɑʔt ‘vein’ ɫsɑt-ɑay ‘veins’ 14. tisuʔx ‘quebracho (tree)’ tisx-uy ‘quebrachos’ 15. k’utsaʔx ‘old man’ k’utsx-as ‘old men’
Set III 16. towɑk ‘river’ towx-ɑy ‘rivers’ 17. ɸinɑk ‘tobacco’ ɸinx-ɑy ‘tobaccos’ 18. ituʔk ‘my arm’ itx-uy ‘my arms’ 19. tsanuʔk ‘duraznillo (tree)’ tsanx-uy ‘duraznillos’ 20. namač ‘axe’ namx-ay ‘axes’ 21. šateč ‘head’ šatx-es ‘heads’ 22. titeč ‘plate’ titx-ey ‘plates’ 23. k’atseč ‘wood chip’ k’atsx-es ‘wood chips’ 24. axpɑyič ‘your house’ axpɑyx-ey ‘your houses’
Reconstructions: 1. *xutsax ‘vulture’ *xutsax-as ‘vultures’ 2. *ɸatsux ‘centipede’ *ɸatsux-us ‘centipedes’ 3. *snomax ‘ash’ *snomax-as ‘ashes’ 4. *ɫtɑsex ‘seed’ *ɫtɑsex-ey ‘seeds’ 5. *kutsxanax ‘thief’ *kutsxanax-as ‘thieves’ 6. *ipɑset ‘my lip’ *ipɑset-es ‘my lips’ 7. *nasuk ‘guayacán (tree)’ *nasuk-uy ‘guayacans’ 8. *ɸaʔayuk ‘algarrobo (tree)’ *ɸaʔayuk-uy ‘algarrobos’ 9. *axɑyuk ‘mistol (tree)’ *axɑyuk-uy ‘mistols’
10. *inkɑʔp ‘year’ *inkɑʔp-es ‘years’ 11. *ɫuʔp ‘nest’ *ɫuʔp-is ‘nests’ 12. *k’utxaʔn ‘thorn’ *k’utxaʔn-is ‘thorns’ 13. *ɫsɑʔt ‘vein’ *ɫsɑʔt-ɑy ‘veins’ 14. *tisuʔx ‘quebracho (tree)’ *tisuʔx-uy ‘quebrachos’ 15. *k’utsaʔx ‘old man’ *k’utsaʔx-as ‘old men’
16. *towɑk ‘river’ *towɑk-ɑy ‘rivers’ 17. *ɸinɑk ‘tobacco’ *ɸinɑk-ɑy ‘tobaccos’ 18. *ituʔk ‘my arm’ *ituʔk-uy ‘my arms’ 19. *tsanuʔk ‘duraznillo (tree)’ *tsanuʔk-uy ‘duraznillos’ 20. *namak ‘axe’ *namak-ay ‘axes’ 21. *šatek ‘head’ *šatek-es ‘heads’
116 22. *titek ‘plate’ *titek-ey ‘plates’ 23. *k’atsek ‘wood chip’ *k’atsek-es ‘wood chips’ 24. *axpɑyik ‘your house’ *axpɑyɑk-ey ‘your houses’
Changes: 1. V > V1 /V1(C)C+__ y, s The vowel of the plural suffix, -Vy or -Vs, assimilates completely to the preceding vowel, as in 1. *xutsax-Vs ‘vultures’ > *xutsax-as (which later becomes xutsxas by other changes). In the reconstructed forms given above, the vowel is given in its already assimilated form, though perhaps *xutsx-Vs would be more representative. [It is not easy to determine what the vowel of the plural suffixes was originally – usually it is a totally assimilated copy of the vowel preceding it in the stem to which the plural suffix is attached, though there are some exceptions, usually with i or e, with suggest perhaps it may have been i or e originally. This, however, is beyond the scope of the data presented here.] 2. V > Ø /VC__(ʔ)C+V The vowel of a stem was lost if followed by a single consonant or a glottal stop and consonant when a vowel-initial suffix is attached so long as it did not result in a root with no vowels (that is, as long as the form to which the suffix was attached had more than one syllable and would not result in a consonant cluster of more than two consonants together), as. in 1. *xutsax-Vs ‘vultures’ > xutsx-as (by 1. and 2.) 3. ʔ > Ø /__C+V Glottal stop before another consonant in roots was lost when a vowel-initial suffix was attached, as is 11. *ɫuʔp-is ‘nests’ > ɫupis. 4. k > x /C__ Here, *k became x when it followed another consonant. This is seen in 16. *towɑk-Vy ‘rivers’ > towɑk-ɑy by 1., then towɑk-ɑy > towkɑy by 2., and then towkɑy > towxɑy by 4. As is seen in the examples, k came to follow another consonant often due to change 2. which deleted an intervening vowel. 5. k > č / i, e, a __ Here, *k became č when followed by non-back vowels (i, e, a) (not o, u, ɑ; note that /a/ is a central, not back vowel, while /ɑ/ is back), as in 21. *šatek ‘head’ > šateč.
Relative chronology: Change 1. (vowel-assimilation) came before change 2. (vowel-loss), since if the vowel were lost in 2. before assimilation in 1., the vowel in the root to which the suffix vowel assimilates by change 2. would not be there to change the suffix vowel. Change 3. (ʔ-deletion) does not interact with the other changes and so could have taken place at any time relative to the others. Change 2. preceded change 4. If the sequence had been the opposite, erroneous forms would have resulted, as for example with 16. *towɑk-Vy ‘rivers’ > towɑk-ɑy by 1., the 4. (k > x before a consonant) would not have been applicable, since k would be between vowels, then later change 2. (vowel-deletion) would give erroneous results, towɑkɑy > ✘towkɑy. Change 4. took place before change 5, changing k to x thus not leaving a k to change to č before i, e, or a, as in 22. *titek-ey ‘plates’ > titkey by 2., at which point if Change 5. had taken place, we would get then titčey, making later change 4. (k > x) inapplicable to this form, since
117 titčey would no longer has a k which could change to x in 4., leaving us with the erroneous form ✘titčey.
Derivations of the words of 12, 14, 18, and 20: 12. *k’utxaʔn ‘thorn’ *k’utxaʔn-is ‘thorns’ 1. -- [exception to change 1.] 2. -- -- [not applicable, clusters of three C’s (txn) not permitted] 3. -- k’utxan-is 4. -- -- 5. -- -- Result k’utxaʔn k’utxanis
14. *tisuʔx ‘quebracho (tree)’ *tisuʔx-Vy ‘quebrachos’ 1. -- tisuʔxuy 2. -- tisʔxuy 3. -- tisxuy 4. -- -- 5. -- --
18. *ituʔk ‘my arm’ *ituʔk-Vy ‘my arms’ 1. -- ituʔkuy 2. -- itʔkuy 3. -- itkuy 4. -- itxuy 5. -- -- Result ituʔk itxuy
20. *namak ‘axe’ *namak-Vy ‘axes’ 1. -- namakay 2. -- namkay 3. -- -- 4. -- namxay 5. namač -- Result namač namxay
118
EXERCISES FOR CHAPTER 9 SEMANTIC CHANGE AND LEXICAL CHANGE
Exercise 9.1 Attempt to find examples of your own of new vocabulary items which represent some of the categories of lexical and semantic change discussed in this chapter. Try to name or identify the categories involved. You can do this by listening for words that you think are new in the speech of your friends and family or by asking others if they can think of any examples. Slang is a fertile area for new vocabulary and semantic shifts.
There is, obviously, no set answer to this exercise. While the exercise calls for making personal observations, you can get ideas of new vocabulary from a number of internet and library sources. For example, a search for “new words (added to the dictionary)” will reveal new words added to dictionaries in particular years and will lead you to examples such as, bloggable, drama queen, facebook, scareware, supersize, etc. Searches, for example, for “new slang words”, “creation of new words”, “new words in English”, and “neologisms” will take you to numerous others.
Exercise 9.2 Lexical change
The following are a few of the many new words (neologisms) that have been added to English recently. Can you determine where these come from, that is, how they came about? What processes of vocabulary creation, semantic change or other kinds of linguistic changes do you think lie behind the creation of these new words? (You may need to look some of these up to find their meanings, or ask your friends who might know what they mean.)
app, buzzword, de-friend / unfriend, emoticon, guesstimate, mouse potato, to google, LOL, tweet, webinar, wiki, WMD.
app: Clipping, from ‘application’. buzzword: Compounding, buzz and word. de-friend / unfriend: Word-formation. emoticon: Blending, emote + icon. guesstimate: Blending, guess + estimate. mouse potato: [someone who spends excessive amounts of time on the computer] Compounding, on analogy to ‘couch potato’. to google: From brand name, Google (and word-formation, noun > verb). LOL: Acronym, from ‘laughing out loud’. tweet: Metaphor, on similarity with tweet ‘sound made by a small bird’. webinar: Blending, web + seminar.
119 wiki: ultimately a loanword, from Hawai‘ian wikiwiki ‘fast, swift’; originally from WikiWikiWeb, which was introduced and named by Ward Cunningham in 1995. WMD: Acronym, from ‘weapons of mass destruction’.
Exercise 9.3 Semantic change
Look up the following words in a dictionary which provides basic etymologies for words. (The Oxford English Dictionary is generally recognized as the primary authority in this area and is recommended here, although a number of other dictionaries also provide useful etymological information.) Determine what change in meaning has taken place in each word. State which type of semantic change is involved (from among the types defined in this chapter). For example, if you were to see villain in the list, you would look it up and find out that it originally meant ‘person of the villa/farm’ but has changed its meaning to ‘criminal, scoundrel’, and you would state that this is an example of degeneration (or pejoration).
corpse; crafty; disease; fame; journey; officious; science; starve; thing; vulgar
corpse: a variant spelling of corps, which is derived from French cors ‘body’, ultimately from Latin corpus ‘body’. The semantic shift is from ‘body’ to ‘dead body’, an example of narrowing. (Some may imagine that degeneration may have been involved, also.) crafty: from Old English cræftig ‘strong, powerful’, expanded to include ‘skillful’. (Based on the root craft, Old English cræft ‘strength, power’ and later ‘skill’.) The semantic shift is from ‘skillful’ to ‘cunning, sly’, an example of degeneration. disease: borrowed from Old French desaise ‘lack, discomfort, trouble, misfortune, sickness’ (derived from des- ‘without’ + aise ‘ease’). The semantic shift from ‘discomfort’ to ‘illness’ is an example of degeneration. It may also be seen perhaps as having involved narrowing, metonymy, or litotes. fame: borrowed from Old French fame ‘fame, reputation, renown, rumor’, from Latin fāma ‘talk, rumor, report, reputation’ (based on an Indo-European root ‘to speak’). The goddess Fāma personified rumor in Roman mythology. The semantic shift from ‘report, talk’ to ‘reputation, renown’ is an example of elevation. (Narrowing is also involved.) journey: borrowed from Old French journée ‘a day’s work, day’s travel’, ultimately from Vulgar Latin diurnum ‘day’. The semantic shift was from ‘a day’s work or travel’ to ‘travel, trip, journey’. It may be seen as an example of metaphor or broadening (from how far one jouneyed in a day to the jouney generally). officious: borrowed from Latin officiōsus ‘dutiful’ (derived from officium ‘duty, service’. The semantic shift from ‘dutiful’ to ‘intrusive, obstructive’ (asserting authority in an annoyingly domineering way about trivial matters). It is an example of degeneration (also involving narrowing). science: borrowed from Latin scientia ‘knowledge’ (from the root scire ‘to know’). The semantic shift from ‘knowledge’ to ‘knowledge acquired by a particular method’ is an example of narrowing.
120 starve: from Old English steorfan ‘to die’ (compare the German cognate sterben ‘to die’). The semantic shift was first from ‘to die’ to ‘to die of cold’ and then to ‘to die of hunger’; it is an example of narrowing. (It might be thought also to involve hyperbole.) thing: from Old English þing ‘governing assembly’. The semantic shift from ‘assembly’ to ‘object, entity, matter’ is an example of metonymy, where people talked about ‘matters, events, entities’ in assemblies. (Some might see synecdoche.) vulgar: borrowed from Latin vulgāris ‘of or pertaining to the common people’ (from the root vulgus ‘the common people, multitude, crowd’). The semantic shift from ‘of the common people’ to ‘lacking sophistication or good taste, unrefined’ and on to include ‘coarse, rude, making explicit and offensive reference to sex or bodily functions’. It is an example of degeneration. (Narrowing can also be said to be involved.)
Exercise 9.4 In the following examples of semantic change, identify the kind of semantic change involved (widening, narrowing, metonymy and so on).
1. Spanish cosa ‘thing’ < Latin causa ‘cause, question, matter’. Metonymy. 2. Spanish dinero ‘money’ < Latin dēnāriu ‘coin (of a particular denomination)’. Widening. 3. Spanish pariente ‘relative’ < Old Spanish pariente ‘parent’. Widening. 4. Spanish segar ‘to reap (to cut grain, grass with a scythe)’ < Latin secāre ‘to cut’. Narrowing. (Perhaps also metonomy or synecdoche.) 5. Old Spanish cuñado ‘relation by marriage’ shifted to ‘brother-in-law’ in Modern Spanish. Narrowing. 6. Mexican Spanish muchacha, formerly only ‘girl’, now has a primary meaning ‘maid, servant woman’ in some contexts. Narrowing. (Perhaps also some degeneration.) 7. Modern Spanish siesta ‘afternoon nap (rest period during the heat of the day)’ < Old Spanish siesta ‘midday heat’ (ultimately from Latin sexta (hora) ‘sixth (hour)’). Metonomy. 8. English gay ‘homosexual’ is the result of a recent semantic shift, where the original sense, ‘cheerful, lively’, has become secondary; the shift to the ‘homosexual’ sense perhaps came through other senses, ‘given to social pleasures, licentious’, which the word had. Views may vary here; some may believe it involves metaphor, some elevation, others degeneration, some taboo replacement or euphemism. 9. English to spill formerly meant (from c.1300 to 1600) ‘to destroy by depriving of life, to put to death, to slay, to kill’. Hyperbole. 10. French cuisse ‘thigh’ < Latin coxa ‘hip’ (Spanish cojo ‘lame, crippled’ is thought also to be from Latin coxa ‘hip’). Synecdoche or metonomy.
121 11. Spanish cadera ‘hip’ < Latin cathedra (from Greek) ‘seat’. Synecdoche or metonomy. 12. Spanish ciruela ‘plum’ < Latin prūna cēreola ‘waxy plum’ (prūna ‘plum’ + cēreola ‘of wax’). Displacement (ellipsis). 13. French viande ‘meat’ formerly meant ‘food’ in general. (This change parallels English meat which originally meant ‘food’.) Narrowing. 14. Spanish depender ‘to depend’ < Latin dēpendere ‘to hang’. Metonomy. 15. English lousy ‘worthless, bad’ < ‘infested with lice’. Metonomy.
122
EXERCISES FOR CHAPTER 11 SYNTACTIC CHANGE
Exercise 11.1 Syntactic change in Panare
Consider the following from Panare (a Cariban language of Venezuela). (NOTE: y = IPA [j], ñ = IPA [ɲ], č = IPA [tʃ].) The basic word order is verb first and subject final.) When the subject is ‘I’ or ‘you’, no copula (form of the verb ‘to be’) is required in the present tense, as in:
(1) maestro yu (2) maestro amən teacher I teacher you ‘I am a teacher’ ‘You are a teacher’
However, with a third-person subject, a copula is obligatory. With an inanimate subject, the copula is mən, as in (3):
(3) eʔčipen mən manko fruit mən mango ‘Mango is a fruit’
For this exercise, such examples with inanimate subjects are not so relevant. However, with an animate subject, the copula is either kəh or nəh, with a difference in meaning. Sentences (4) and (5) show that sentences with third-person subjects but with no copula are ungrammatical (here /✘/ means ungrammatical):
(4) ✘maestro eʔñapa (5) ✘eʔčipen manko teacher Panare fruit mango
Sentences (6) (7) and (8) illustrate the kəh and nəh copulas and their difference: (6) maestro kəh eʔñapa (7) eʔčipen mən manko teacher kəh Panare fruit mən mango ‘This Panare here is a teacher’ ‘Mango is a fruit’
(8) maestro nəh eʔñapa teacher nəh Panare ‘That Panare there is a teacher’
Now consider some demonstratives. The demonstratives məh ‘this person whom I can see now’ and kən ‘that person whom I can’t see now’ at first glance appear to behave straightforwardly, as in (9) and (10):
(9) maestro kəh məh (10) maestro nəh kən teacher kəh this.guy teacher nəh that.guy ‘This guy is a teacher here’ ‘That guy is a teacher there’
But consider the additional Panare copular sentences in (11) through (14) (note here that /y/ changes to
123 /č/ after /h/, so that yu ‘I’ in this example is ču in this context):
(11) maestro nəh məh (12) maestro nəh ču teacher nəh this.guy teacher nəh I ‘This guy was a teacher’ ‘I was a teacher’
(13) maestro nəh amən (14) maestro kəh kən teacher nəh you teacher kəh that.guy ‘You were a teacher’ ‘That guy is being a teacher right now’ (that is, he is off somewhere performing his teaching duties at this very moment). Though originally not possible, notice also that kəh and nəh now can also occur with ordinary verbs, as in (15) through (18) (the question mark indicates a sentence which sounds very strange to native speakers):
(15) əʔ púmanəpəh kəh Toman (16) ? əʔ púmanəpəh nəh Toman be falling kəh Thomas be falling nəh Thomas ‘Tom is falling’ ‘Tom is falling (but I can’t see him)’ (17) yɨupúmən kəh Toman (18) yɨupúmən nəh Toman fall kəh Thomas fall nəh Thomas ‘Tom is going to fall!’ ‘Tom is going to fall one day’, or ‘Tom fell’ State the syntactic changes which have affected kəh and nəh. Explain the historical development of these items as best you can using the terms and mechanisms presented in this chapter. (Based on Gildea 1993, exercise prepared by Verónica Grondona.)
State the syntactic changes which have affected kǝh and nǝh. Stage I: nǝh Copula with human visible non-pronominal subject (visible, present, or near) kǝh Copula with human non-visible non-pronominal subject (not visible, not present, or distant).
Stage II: Reanalysis: nǝh and kǝh copulas are reanalyzed also as verbal tense-aspect markers: nǝh takes on the added function of ‘past’ (non-visible) verbal tense-aspect marker kǝh takes on the added function of ‘non-past’ (visible) verbal tense-aspect marker.
Exercise 11.2 Syntactic change in Estonian
Compare the sentences in this exercise, which represent different stages of Estonian (a Finno-Ugric language); explain what changed and identify the kinds of changes or the mechanisms involved.
Stage I: Estonian had two alternative constructions for subordinate clauses involving the complements of speech-act and mental-state main verbs, illustrated in (1) and (2) (Gen = ‘genitive’, Nom = ‘nominative’, Part = ‘participle’, Pres = ‘present indicative’):
124
(1) sai kuulda, et seal üks mees ela-b got to.hear that there one.NOM man.NOM live-3rd.PRES ‘He/she came to hear that a man lives there’
(2) sai kuulda seal ühe mehe ela-vat got to.hear there one.GEN man.GEN live-PARTICIPLE (same meaning as (1))
Stage II: (1) and (2) remain possible, but the construction in (3) also became possible (note that ‘participle’ became ‘indirect’):
(3) sai kuulda, (et) seal üks mees ela-vat got to.hear (that) there one.NOM man. NOM live-INDIRECT (3a) ‘he/she came to hear that they say a man lives there’/ (3b) ‘he/she came to hear that reportedly a man lives there’
Stage III: (1), (2) and (3) are all possible now, but forms formerly found only in subordinate clauses, as in (3), came to be found also in main clauses, as in (4): (4) ta tege-vat töö-d he.NOM do-INDIRECT work-PARTICIPLE ‘They say he is working’ / ‘Reportedly he is working’
Explain what changed and identify the kinds of changes or the mechanisms involved. Changes: Reanalysis. In Stage II, the Participle -vat was reanalyzed as ‘Indirect’ (reported speech). Extension: In Stage III, the reanalyzed -vat ‘Indirect’ was extended so that it was no longer restricted to subordinate clauses, as in (3) in Stage II, but now can appear also in main clauses, as in (4) in Stage III.
Exercise 11.3 The development of the perfect auxiliaries in Spanish
In the following, the stages in the development of perfect auxiliaries in Spanish from their Latin origins are described and illustrated. On the basis of this information, compare the stages and attempt to determine the changes which took place and to identify the kinds of changes or the mechanisms involved. (Fem = ‘feminine’, Masc = ‘masculine’, Part = ‘participle’, Pl = ‘plural’, PPP = ‘past passive participle’). Stage I, the beginning state for this problem (no changes yet): Latin used expressions with ‘past passive participle’ (PPP) in combination with the verbs tenēre ‘hold’, habēre ‘keep, hold’, and other verbs meaning ‘hold, possess, own’, to represent something as ready or kept in a completed condition, as in (1):
(1) Metuō enim nē ibi vos habeam fatigā-tō-s (Late Latin) fear.I truly lest there you have.I fatigue-PPP.Masc-Pl ‘I fear that I have you tired’ / ‘that I have tired you’ /
125 ‘that you are tired’
This construction with ‘past passive participle’ was quite limited in its occurrence in Classical Latin, but became associated with ‘perfect’ aspect in combination with the development of habēre as an auxiliary. Originally this construction had habēre ‘keep, hold, have’ (a main verb) with the ‘past passive participle’ form as an adjective which modified the direct object (both the logical and surface object) of this main verb (habēre), which agreed in number and gender with this object as its head, as seen in the contrast between (2a) and (2b):
(2a) [habe-ō] [litter-ā-s scrip-t-ā-s] have-I letter-Fem-Pl.Acc write-PPP-Fem-Pl.Acc ‘I have written letters’ (‘I have letters that are written’) (2b) [habe-ō] [scrip-tum libr-um] have-I write-PPP.Masc.Acc.Sg book-Masc.Acc.Sg
As seen here, the Past Passive Participles scriptās in (2a) and scriptum in (2b) are declined as adjectives that agree in case, gender, and number with the head nouns, litterās in the (2a) and librum in (2b); the Past Passive Participle is not part of the main verb (habeō ‘I have, hold’ here).
Stage II: In Old Spanish, haber (spelled aver in Old Spanish, from Latin habēre ‘to have, hold’) in such constructions began to lose its possessive meaning and to consolidate the auxiliary function, resulting in compound tenses, but still with agreement in gender and number between the participle and the direct object until the mid-sixteenth century, as illustrated in (3) (where the -o-s ‘masculine plural’ of hechos ‘made’ agrees with the -o-s ‘masculine plural’ of enemigos ‘enemies’):
(3) Los había ... he-ch-o-s enemig-o-s de estotros Them had make-Past.Part-Masc-Pl enemy-Masc-Pl of these.others ‘He had made enemies of these others’ (from Hernán Cortés)
Stage III: Gradually, the haber + PPP construction changed, eliminating the requirement that ‘past passive participle’ must agree in number and gender with the noun which it modified, losing its passive sense, with the verb haber becoming the ‘perfect auxiliary’, and Modern Spanish no longer permits agreement between the participle and the object, as in (4):
(4) Hemos escri-to cart-a-s have.we write-Past.Part letter-Fem-Pl ‘We have written letters’
The adjectival participle source with number and gender agreement still survives in other contexts (but not in the perfect construction with forms of the verb haber), for example:
(5) Tenemos cart-a-s escri-t-a-s en tint-a roj-a have.we letter-Fem-Pl write-Past.Part-Fem-Pl in ink-Fem.Sg red-Fem.Sg ‘We have letters written in red ink’.
126
In the series of changes described here, the meaning is no longer ‘X possesses that which has been done’, but ‘X has done’, and is accompanied by the structural change of haber from main verb to an auxiliary.
Stage IV: Additional changes in connection with the new ‘perfect’ construction also came about. First, the verb ser ‘to be’ had formerly also been an auxiliary used with certain intransitive verbs (especially verbs of motion) (as in (6a) and (7a)), but this was replaced by the auxiliary haber, as seen in the Modern Spanish equivalents in (6b) and (7b):
(6a) Old Spanish ella es naci-d-a she is born-Past.Part-Fem (6b) Modern Spanish ella ha naci-d-o she has born-Past.Part ‘she has been born’ (7a) Old Spanish ellos son i-d-o-s they are go-Past.Part-Masc-Pl (7b) Modern Spanish ellos han i-d-o they have go-Past.Part ‘they have gone’ (Lapesa 1981: 212)
Second, the word order changed, placing the participle closer to the auxiliary, for example from the equivalent of ‘I have a letter written’ (as in (2)) to ‘I have written a letter’ (as in (4)).
Identify the kinds of changes or the mechanisms involved. In Stage I there are no changes, though some changes had taken place earlier to get to Stage I: Latin came to use expressions with ‘past passive participle’ (PPP) in combination with the verbs tenēre ‘hold’, habēre ‘keep, hold’, and other verbs meaning ‘hold, possess, own’, to represent something as ready or kept in a completed condition. The verb habēre developed to become an auxiliary, a grammaticalization of habēre ‘keep, hold, have’ (main verb) to auxiliary. Changes: In Stage II, there was (1) reanalysis (grammaticalization) of haber as an auxiliary verb (with the subject of inflected haber as also the subject of the Participle, and (2) bleaching of haver of its possessive meaning. The new construction co-existed for a while with original one as in Stage I. In Stage III: reanalysis. In the series of changes described in this stage, the meaning is no longer ‘X possesses that which has been done’, but ‘X has done’, and is accompanied by the completed structural change of haber from main verb to an auxiliary. There was extension in that Modern Spanish no longer permits the agreement between the participle and the object that was required earlier, and after the change the former object of the participle is treated as just an ordinary object of the verb phrase. In Stage IV: there was extension of haber to all intransitive verbs where previously intransitives of verbs of motion took only ser ‘to be’ as their auxiliary.
127 Exercise 11.4 Finding examples of grammaticalization
The following are some of the most common pathways of grammaticalization (that is, lexical sources which often become grammatical morphemes as a result of grammaticalization changes). Attempt to find examples from English (or from other languages you may be familiar with) which might illustrate these processes. (A few which are extremely common around the world are also included even though English alone may not offer examples.) As an example, for ‘go to’ > FUTURE, you might list English ‘going to’ > FUTURE as in ‘Hermione is going to marry Ron.’
(1) Allative (‘to’) > complementizer (for example, marker of infinitives) (2) ‘come’ > future (3) Copula (‘to be’) > obligation (such as ‘must’, ‘should’) (4) Demonstrative pronoun (such as ‘this’, ‘that’) > definite article (such as ‘the’) (5) ‘get’ > passive (6) ‘have’ (possession) > obligation (7) ‘have’ (possession) > perfect or completive aspect (8) ‘keep’ (‘hold’, ‘grasp’) > continuous (9) ‘keep’ > possession (‘have’) (10) ‘man’ > indefinite pronoun (11) ‘need’ > obligation (12) ‘one’ > indefinite pronoun, indefinite article (13) ‘owe’ > obligation (14) ‘say’ > quotative (15) ‘say’ > conditional (16) ‘want’ > future (17) Wh-question word (such as ‘what?’, ‘which?’, ‘who’) > relative pronoun, relative clause marker.
There are no set answers to the request to find examples that match the common pathways of grammaticalization listed here. You will need to think of examples on your own. You can check various publications and websites, if you are unable to find appropriate cases on your own. There are numerous books and articles on grammaticalization with abundant examples which should give you ideas of how to find your own examples.
128
EXERCISES FOR CHAPTER 14 DISTANT GENETIC RELATIONSHIP
Exercise 14.1 The Xinka-Lenca hypothesis
Handbooks and encyclopedias continue to report that Xinca (Xinka) and Lenca are related to one another and belong in a single family group. Essentially all the evidence ever published for this hypothesis is presented below. Evaluate this proposed evidence based on the criteria and considerations discussed in this chapter. What do you conclude about the strength of the supporting evidence? Background. Xinka is actually a family, now called Xinkan, of four languages in southeastern Guatemala, two extinct and the two with only a couple of semi-speakers surviving. Lenca is also a small family, Lencan, of two languages, Chilanga (Salvadoran Lenca) and Honduran Lenca, both recently extinct. Walter Lehmann (1920:767) suggested the hypothesis that links the two families, though he also included other languages in his comparison, a fact now long forgotten. Lehmann’s evidence for the proposed relationship is reproduced here (with the form from Guazacapán Xinka in parentheses). Observations that may be relevant in your deliberations are also included as notes below the forms. Xinka Lenca 1 ‘one’ ical (ik’aɬ) etta, ita 2 ‘two’ bi-al, pi-ar, pi (piʔ) pe 3 ‘three’ vuaal-al, hual-ar (waɬ, waɬa) laagua, lagua 4 ‘four’ iri-ar ((h)irha) heria, erio (also sa, aria, eslea) NOTE: words for numbers higher that ‘two’ are widely borrowed in languages of this part of Central America; these include forms similar to those for ‘two’, ‘three’, and ‘four’ in these languages. 5 ‘water’ uÿ (u:y) cuy (invierno [‘winter’]) 6 ‘night’ suma (sa-sɨ’ma ‘in the dark’) ts’ub (Nacht [‘night’]) 7 ‘dark, black’ ts’ama (sɨ’ma) ts’ana-uamba (Morgen (grauen) ‘morning (to dawn)’) 8 ‘shade’ ti-tzuma (ti-sɨ’ma ‘in the dark’) saba NOTE: The Xinkan forms in 5, 6, and 7 all involve the same root, /sɨ’ma/ ‘dark, black’. 9 ‘dog’ xusu (
129 Evaluate this proposed evidence: Xinca Lenca 1 ‘one’ ical (ik’aɬ) etta, ita These are not phonetically very similar; the sounds do not correspond systematically (neither k’ : t nor lh : Ø is a recurring sound correspondence).
2 ‘two’ bi-al, pi-ar, pi (piʔ) pe Short forms.
3 ‘three’ vuaal-al, hual-ar (waɬ, waɬa) laagua, lagua 4 ‘four’ iri-ar ((h)irha) heria, erio (also sa, aria, eslea) NOTE: words for numbers higher that ‘two’ are widely borrowed in languages of this part of Central America; these include forms similar to those for ‘two’, ‘three’, and ‘four’ in these languages. The forms in 3 and 4 involve borrowing.
5 ‘water’ uÿ (u:y) cuy (invierno [‘winter’]) Semantic latitude: water / winter Short forms. No explanation for why Ø :
6 ‘night’ suma (sa-sɨ’ma ‘in the dark’) ts’ub (Nacht [‘night’]) 7 ‘dark, black’ ts’ama (sɨ’ma) ts’ana-uamba (Morgen (grauen) ‘morning (to dawn)’) 8 ‘shade’ ti-tzuma (ti-sɨ’ma ‘in the dark’) saba NOTE: The Xinka forms in 5, 6, and 7 all involve a single root, sɨ’ma ‘dark, black’. Since the Xinka forms in 6, 7, and 8 are all based on a single root, sɨ’ma ‘dark’, there are not three separate and independent cognate sets as evidence for the hypothesis, but at best only one. This single Xinkan root cannot be cognate with three different words in Lenca (unless they, too, are derived from a single root).
9 ‘dog’ xusu (
10 ‘cough’ ojo [
11 ‘maize’ au, aima (ayma) ama, aima NOTE: forms similar to aima, ama, eima for ‘maize’ are found widely in other languages of this region; they involve borrowings. Borrowing.
130 12 ‘bean’ xinak (ṣ̌i’nak) shinag NOTE: Mayan languages border both Xinkan and Lencan territory. Cf. Cholan-Tzeltalan (Mayan) *čenek’ ‘bean’, from Proto-Mayan *kinaq’ ‘bean’. Terms for ‘bean’ are borrowed from Mayan also in some other languages of this part of Central America. Borrowings.
What do you conclude about the strength of the supporting evidence? With only 12 comparisons to start with, this is an extremely small amount of evidence upon which to build a hypothesis of genetic relationship, but since nearly all 12 of the comparisons that were offered are seriously challenged for various reasons, the Xinca-Lenca hypothesis should be abandoned.
Exercise 14.2 An “Amerind” putative “etymology.”
Greenberg and Ruhlen (1992) presented the forms given here as one of the strongest examples of a putative cognate set to support the “Amerind” hypothesis. Evaluate these data on the basis of the criteria and considerations discussed in this chapter. Background: Greenberg’s (1987) “Amerind” hypothesis would group all the language families and isolates of the Americas except Eskimo-Aleut and Na-Dene. Most specialists in American Indian linguistics believe there are about 180 independent language families and isolates, not just one big one – that is, they believe the evidence available today is insufficient to reduce this number of families by much, though it may be possible that in the remote past they were related, just so long ago we can no longer demonstrate it.
Nootka t’an’a ‘child’ Yuchi tane ‘brother’ Totonac t’ána-t ‘grandchild’ Coahuilteco t’an-pam ‘child’ Proto-Uto-Aztecan *tana ‘daughter, son’ Miskito tuk-tan ‘child, boy’ Warrau [Warao] dani- ‘mother’s sister’ Aymara tayna ‘firstborn child’ Masaca tani-mai ‘younger sister’ Urubu-Kaapor ta’ïn ‘child’ Pavishana tane ‘my son’ Lengua tawin ‘grandchild’ Tibagi tog-tan ‘girl’ Yurok tsin ‘young man’ Mohawk -’tsin ‘male, boy’ Molale pēn-t’in ‘my elder brother’ Yana t’inī-si ‘child, son, daughter’ Cuicatec ’díínó ‘brother’ Changuenga sin ‘brother’ Millcayac tzhœng ‘son’ Tehuelche den ‘brother’ Tiquie ten ‘son’
131 Mocochi tin-gwa ‘son, boy’ Yagua dēnu ‘male child’ Tacana u-tse-kwa ‘grandchild’ Guato china ‘older brother’ Coeur d’Alene tune ‘niece’ Yuchi ts’one ‘daughter, son’ Central Sierra Miwok tūne- ‘daughter’ Salinan a-t’on ‘younger sister’ Taos -t’út’ina ‘older sister’ Cayapa tsuh-ki ‘sister’ Tehuelche thaun ‘sister’ Tiquie ton ‘daughter’ Morotoko a-tune-sas ‘girl’ Nonuya -tona ‘sister’ Tacana -tóna ‘younger sister’ Pikobyé a-ton-kä ‘younger sister’
Problems: Semantics: almost any kin term involving the same or younger generation or any term for a younger male or female would fit. It is far too easy to find phonetically similar forms in languages around the world that fit this broad semantic range. Phonology: apparently the target is roughly some alveolar-like sound, followed by any vowel, followed by an n-like sound (which is not always present), so, something like: t’/t/d/ts/s/tzh/ch/th-V(-n/ng/h/Ø). It is far too easy to find forms that fit this range of meanings and phonetic variation across languages by accident – a few that come to mind are: English: son, tot, teen German: Tante ‘aunt’ Spanish: china ‘girl, young women’ Maori: tāina ‘younger brothers (of a male), younger sisters (of a female)’, tēina ‘younger brothers (of a male), younger sisters (of a female)’.
Exercise 14.3 Macro-Panoan distant genetic relationship
All of the evidence presented by Joseph H. Greenberg (1987:74-78) for his proposed Macro- Panoan hypothesis (part of his larger Amerind proposal) is repeated here. Evaluate it based on the criteria and considerations in this chapter. What other possible explanations (other than that of cognates inherited from some common ancestor) do you see for some of the similarities among the lexical items compared from the different languages? What kinds of problems do you notice? List the problems involved in each lexical set. (Pay attention also to the number of languages from which potential evidence is cited in comparison to the total number of languages hypothesized to belong to this group.) After you have set aside forms that potentially have other, non-genetic explanations, what evidence (if any) do you find that might support a possible genetic relationship among the languages compared here? Greenberg’s Macro-Panoan hypothesis would group several South American language families and isolates: Panoan, Tacanan (recognized as Pano-Tacanan), Moseten [Mosetenan],
132 Mataco[an]], Guaicuru[an], Charruan, Lule, Vilela (recognized as Lule-Vilela), and Mascoy[an]. Note that though separated by Greenberg, Panoan and Tacanan belong to the Pano- Tacanan family. Clarification of some of the names and some of the forms not included in Greenberg’s data are given in square brackets ([ ]).
1 BE ABLE [Mascoyan:] Lengua wan(-či), wan(-kje). Mataco[an]: Chulupi [Nivaclé] ha- wanaia [no such form exists; xa- ‘1st person’]. Problems: Only 2 languages are compared from many in the set of languages involved. Spurious form.
2 ANIMAL Guaicuru[an]: Toba-Guazu sigiak. [Mascoyan:] Lengua askok. Mataco[an]: Vejoz [Wichí] ɬokue [no such word exists in Wichi; ɬokwe is ‘jug’; lo is the classifier for possessed domestic animals and may be involved here]. Problems: spurious form.
3 ANSWER (v.) Mataco[an]: Choroti kamtini ‘speak’. Panoan: Cazinaua køma. Cavineña kiema. Problems: semantics: answer / speak. Only two languages compared.
4 ANUS Guaicuru[an]: Caduveo -auio ‘buttocks’. Mataco[an]: Choroti i-we, Vejoz [Wichí] wex [‘tail, backside’]. Moseten jive ‘buttocks, anus.’ Panoan [Pano-Tacanan]: Caripuna wahaa ‘open’. Tacanan [Pano-Tacanan]: Huarayo wexa ‘opening,’ Chama wexa ‘hole’. Problems: semantics: buttocks, anus / tail / opening / opening / hole. Short: perhaps only w/v match across the proposed cognates.
5 AWAKE Charruan: Chana inambi. Guaicuru[an]: Toba-Guazu tom ‘awake, dawn.’ Mataco[an]: Vejoz [Wichí]: nom (intransitive) [n-om ‘come, arrive’, n- ‘directional’]. Panoan: Proto-Panoan (S) *nama ‘to dream.’ Problems: semantics: awake, dawn / come, arrive / dream. Morphological analysis (n-om).
6 BACK [Mascoyan:] Lengua ak-puk, (eja-)puk ‘behind.’ Panoan: Shipibo puika. Tacanan: Cavineña ebekakwa, Chama kiibaaxaxe ‘behind.’ Semantics?: back / behind.
7 BAD Guaicuru[an]: Guachi [Guachi is not a Guaicuruan language, though there is a hypothesis that it may be related]
8 BAT Guaicuru [?]: kahit (h < s). Panoan: Proto-Panoan (S) *kaši.
133 Problems: Only two languages are compared from many in the set of languages involved.
9 BE Lengua [Mascoyan]: Mascoy h-. [Matacoan:] Mataco [Wichí] ihi, hi [i- ‘to be’, i- hi ‘be-LOCATOVE’]. Problems: short forms. Phonology: perhaps nothing matches (h : i)? Only two languages are compared from many in the set of languages involved.
10 BEAR (v.) Guaicuru[an]: Mocovi koo, Toba-Guazu koe. Lule kaa ‘born.’ [Matacoan:] Mataco [Wichí] ko, Vejoz [Wichí] ko. Panoan: Proto-Panona (S) *kai ‘to bear, mother,’ Chacobo ko ‘born.’ Tacanan: Chama kwaja ‘be born.’ Problems: short forms.
11 BEFORE Lengua [Mascoyan]: Lengua, Mascoy nanič, Lengua nahno, nahtu ‘mucho anteo’ [anteo is unclear in Spanish]. Mataco[an]: Chulupi [Nivaclé] naxeš ‘forward’ [no such form exists, possibly from nɑx- ‘to end, terminate’; probably a mixture of nɑyiš ‘road’ and the verb derived from it, nɑyi-n ‘to anticipate, prepare, be first, go on ahead’, where one translation in Spanish is adelantarse ‘to go ahead, to go forward’, which is similar to adelante ‘ahead, forward, in front of, before’], Payagua [not Matacoan, sometimes hypothesized as belonging to Guaicuruan or to proposed Macro-Guaicuruan with Guaicuruan, Matacoan, Payaga, Guachi, but not demonstrated] inahi. Moseten
12 BLOOD Guaicuruan: Toba t-auo, Lule ewe. Mataco[an]: Chunupi [Nivaclé] woi [woʔy]. Tacanan: Chama woʔo ‘red.’ Problems: semantics: blood / red. Phonology: only w matches.
13 BODY Lule toip [-p ‘3rd Pers Possessive Pronoun’]. Mataco[an:] Mataco [Wichí]: tape [t- ‘3rd Pers Possessive Pronoun’]. Tacanan: Cavineña etibo ‘trunk’ [e-tibu: e- ‘Pronoun’, tibu ‘base’]. (Lule toip, however, may contain a suffix -p.) Problems: semantics: body / base (mistaken gloss of ‘trunk’ instead of ‘base’). Morphological analysis: it is slightly fudged: Lule toi ‘body’ + -p ‘3rd Pers Possessive Pronoun’. Notice that the -p is segmented in all the other Lule forms cited which end in p. Phonology: with -p segmented off the Lule form, only t matches in these compared forms.
14 BREAK Lengua [Mascoyan]: Mascoy pok- (intransitive). Mataco[an:] Mataco [Wichí] puhwoje [pux-u ‘break, explode’], Suhin [Nivaclé] poktoče (intransitive) [pɑkxet-ši: pɑkxet ‘break’ + -ši ‘indefinite direction or location’]. Moseten fok. Problems: Onomatopoeic.
15 BREAST Lengua [Mascoyan]: Lengua namakuk, Kaskiha neme ‘nipple’. Lule ineme ‘milk’. Problems: semantics: breast / milk / nipple.
134 Onomatopoeic (cf. nursing, suckling noises). Only two families are compared (two languages from one, one language from the other).
16 BROTHER Charruan: Charrua inčala. Lule kani ‘younger brother.’ Mataco[an:] Mataco [Wichí] čila ‘older brother’ [kyila], činix ‘younger brother’ [kyinix], Choroti kiili ‘older brother,’ kiini ‘younger brother.’ Vilela ikelebepe. (Perhaps two related roots for older and younger brother [definitely two different lexical items in the Matacoan languages cited].) Problems: two separate Matacoan cognate sets are conflated into one here.
17 CLOSE (v.) Mataco[an:] Choroti pone, pione ‘close, cover,’ Vejoz [Wichí] ponhi ‘imprison,’ Towothli [Maká] aponik ‘cover.’ Tacanan: Cavineña pene [‘cover, protect’]. Problems: Forms from only Matacoan languages and one Tacanan language are compared. Semantics: close, cover / imprison / cover, protect.
s 18 COLD1 [Mascoyan]: Lengua math(-kaiyi) ‘be cold.’ Panoan: Proto-Panona (S) *mat i ‘be cold.’ Problems: only two languages are compared.
19 COLD2 Lule kei. Mataco[an:] Enimaga [Maká] koija, Chunupi [Nivaclé] kui [k’uy]. Problems: only Forms from two groups are compared, Lule and two Matacoan languages.
20 CUT Lengua [Mascoyan]: Guana čečet ‘cut up.’ Mataco[an]: Suhin [Nivaclé] siči [note Nivaclé has 10 distinct verbs which translate ‘to cut’; the one intended is probably seʔx ‘to cut up’, perhaps seʔx-ši ‘cut up-indefinite location or direction’], Choroti esita, ešita. Panoan: Proto-Panona (S) *ṣ̌aʔtɨ. Problems: Phonology: only sounds representing one consonant in the compared forms (č / s / ṣ̌ ) match across the set. The mistaken form of Suhin makes it appear more similar than it is. Spurious (?) or at least erroneous Suhin [Nivaclé] example.
21 DARK Guaicuru[an]: Toba, Mocovi epe, pe ‘night.’ Mataco[an]: Choroti pe ‘shadow.’ Tacanan: Chama kea-apo ‘night,’ Tacana apu- ‘dark.’ Problems: Short forms (only p matches). Semantics: night / shadow /dark.
22 DIG Mataco[an]: Vejoz [Wichí] tih, Mataco [Wichí] tiho [tix-i ‘dig’]. Tacanan: Chama teo. Problems: Only two languages – Wichí and Chama – are compared. Phonology: apparently only t is intended to match.
23 DOG Mataco[an]: Suhin [Nivaclé] nuu, Choropi [Nivaclé] nuux. Panoan: Proto- Panoan (S) *ʔino, *ʔinaka. Problems: Only two “languages” are compared, Nivaclé and Proto-Panoan. Phonology: Only n matches; perhaps u / o are intended to match, also.
135 24 DOOR Lule atsiki-
25 DRESS (v.) Lule tala ‘clothing,’ talaks. Mataco[an:] Mataco [Wichí] tula ‘clothing’ [form unknown]. Problems: Only two languages are compared. Unidentifiable form (spurious?).
26 DRY [Mascoyan:] Lengua jima(-gjaji) ‘be dry.’ Mataco[an:] Mataco [Wichí] jɨm ‘dry up’ [Wichí has no ɨ], Suhin, Chuluipi [Nivaclé] jim, Macca [Maká] iim. Moseten jiñ ‘bone.’ Problems: semantics: dry / bone.
27 EMPTY Lule em-p. Mataco[an:] Mataco [Wichí] jim, Chulupi [Nivaclé] jimši [yim-ši ‘to dry up, to end’, yim ‘dry’ -ši ‘indefinite direction or location’. Same root as in 26]. Problems: The same root is compared to different forms in 26 and 27 as though they involved two cognate sets instead of one. Only two groups compared, Lule and Matacoan (two languages). Phonology: short forms; perhaps only m is intended to match (or perhaps also e / i).
28 FEAR1 (v.) Guaicuru[an]: Toba-Guazu nahi. Mataco[an]: Vejoz [Wichí] nowai [the root is oway, n- ‘Middle Voice marker’]. Moseten nojii ‘frighten.’ Panoan: Cashibo noo ‘frighten,’ Nocaman no ‘enemy,’ Panobo, Shipibo nawa ‘enemy.’ Problems: semantics: fear / frighten / enemy. Morphological analysis: with n- of Wichí segmented off, the root is no longer similar. Short form (e.g. no).
w 29 FEAR2 (v.) Lule lako ‘be ashamed.’ Panoan: Proto-Panoan (S) *rak ɨ. Problems: semantics: be ashamed / fear. Only two “languages” are compared.
30 FINISH Lule tum-p ‘be finished.’ Mataco[an]: Choroti temi, Suhin [Nivaclé] timš [im ‘to end, run out’; perhaps based on xa-t-im-ši ‘1stPers-VERB.CLASS-end- INDEFINITE.DIRECTION/LOCATION]. Cf. Tacanan: Cavineña tupu ‘enough.’ Problems: semantics: be finished / end, run out / enough. Morphological analysis: with -t- segmented from the Nivaclé root im, little similarity is left.
31 FLY (v.) Moseten naj. Panoan: Proto-Panoan (S) *noja. Problems: Only two “languages” are compared.
32 GREEN Lule
136 Only two entities are compared, Mosetén and Pano-Tacanan.
33 HANG Moseten pina ‘hammock.’ Panoan: Conibo panea ‘be hung,’ pani ‘hang up,’ Shipibo panni ‘hang up.’ Problems: semantics: hammock / be hung / hang up. Only two entities are compared, Mosetén and Panoan. ‘Hammock’ is a culture item, subject to possible borrowing.
34 HATE Guaicuru[an]: Abipone n-paak ‘hated.’ Moseten fakoj, fakin ‘be angry.’ Problems: semanatics: hated / be angry. Only two languages are compared.
35 HORN Lengua [Mascoyan]: Guana taša. Mosten daš
36 KNEAD [Matacoan:] Mataco [Wichí] pʔon. Moseten puñe ‘knead, mud.’ Problems: only two languages are compared.
37 KNOW Mataco[an]: Vejoz [Wichí] hanex [han- ‘to know’, -ex ‘Applicative’], Choroti hane ‘know, be able’. Moseten (am)-xeñ ‘(no se puede’ [it is not possible]). Panoan: Proto- Panoan (S) *onã ‘know, be able’, Shipibo huna. Problems: Semantics?: ‘to know (be able)’ / ‘it is not possible’
38 LEAF Guaicuru[an]: Toba: l-awe. [Mascoyan:] Lengua wa. Problems: Short forms. Only two languages are compared.
39 LEAVE (ABANDON) Guaicuru[an]: Toba-Guazu jane. Lengua [Mascoyan]: Mascoy jiño. Panoan: Proto-Panoan (S) *ɨnɨ. Vilela jane. Problems: Phonological match?, just n / ñ (or also j / j /Ø / j ?).
40 LOOK Charruan: Chana sola. Guaicuru[an]: Pilaga čelage, Toba-Guazu silaha. Problems: only two entities are compared (Chana and two Guaicuruan languages).
41 LOSE Moseten moñi ‘perish, lose, err.’ Panoan: Cashibo mano ‘forget,’ Cashinahua manu ‘miss.’ Tacanan: Protoo-Tacanan *manu ‘die.’ Problems: semantics: perish, lose, err / forget / miss / die.
42 MAKE Guaicuru[an]: Toba-Guazu uo. Panoan: Proto-Panoan (S) *wa, *ʔa. Tacanan: Proto-Tacanan *a ‘make, say.’ Problems: Short forms. Only two entities are compared, Toba-Guazu and Pano-Tacanan.
43 MANY Guaicuru[an]: Toba-Guazu lamai. [Mascoyan:] Lengua ɬamo. Mataco[an]: Payagua [not a Matacoan language] lehmi ‘all.’ Problems: semantics? many / all.
137 Classification problem: Payagua.
44 MEAT Guaicuru[an]: Pilaga niiak ‘fish.’ [Mascoyan:] Lengua nohak ‘wild animal’. Tacanan: Chama noe, Tiatinagua, Huarayo noči. Vilela nuhu ‘fish.’ Problems: semantics: fish / wild animal. Phonology: perhaps only n matches?
45 MOSQUITO Lengua [Mascoyan]: Mascoy p-aija. Mataco[an]: Choroti eji, Suhin [Nivaclé] iya [(y)iyaʔ]. Problems: Onomatopoeic. Only two entities are compared (Mascoy and two Matacoan languages). Short forms.
46 MOTHER Mataco[an]: Macca [Maká] nana. Tacanan: Proto-Tacanan nene ‘aunt’. Vilela nane. Problems: Nursery forms.
47 MOUSE Guaicuru[an]: Toba-Guazu mekahi ‘bat.’ Moseten meče ‘rat.’ Panoan: Proto- Panoan (S) *maka ‘rat, mouse.’ Cf. Mataco[an]: Mataco [Wichí], Suhin [Nivaclé], Chulupi [Nivaclé] ama, Vejoz [Wichí] ma. Problems: semantics: bat / rat / mouse. (Cf. Short forms?)
s 48 NECK1 Moseten tet
49 NECK2 Lule u(-p). Mataco[an]: Mataco [Wichí], Choroti, etc. wo. Moseten
50 OLD Guaicuru[an]: Guachi [Guachi is not a Guaicuruan language] seera. Mataco[an]: Payagua [Payagua is not a Matacoan language] aheri ‘old woman’. Panoan: Proto-Panoan (S) *ṣ̌ɨnɨ. Tacanan: Proto-Tacanan *ziri. Problems: semantics?: old / old woman. Classification problem: Guachi and Payagua are language isolates, meaning that few comparisons from these two are made with other languages in the set.
51 RED Guaicuru[an]: Toba, Mocovi tok. [Mascoyan:] Lengua eteig-ma. Mataco[an]: Macca [Maká] tek ‘blood’ [not such form exists in Maká; see -athits ‘blood’, -atxuʔ ‘to bleed’, siyixiʔ ‘red’]. Tacanan: Proto-Tacanan *čiaka. Problems: semantics: red / blood.
52 RIB Guaicuru[an]: Mocovi
138 53 ROTTEN [Mascoyan:] Lengua abik. Lule poko ‘to rot.’ Moseten fokoi. Problems: Only two languages are compared.
54 SHOUT Lule se ‘cry.’ Panoan: Shipibo sei, Conibo sije, Cashinahua sa. Tacanan: Proto- Tacanan *ṭsea. Problems: Only two entities are compared, Lule and Pano-Tacanan. Short forms. Semantics(?): cry / shout.
55 SIDE Guaicuru[an]: Toba-Guazu ai, aji, Mocovi ai ‘side,’ Abipone uii. Lule je. Problems: Only two entities are compared, Guaicuruan and Lule. Short forms.
56 SMALL Lengua [Mascoyan]: Mascoy etkok. Mataco[an]: Churupi [Nivaclé] tikin [tik’in], Suhin [Nivaclé] tika [no such form exists, perhaps a mistake for tik’in]. Towothli [Maká[ taake ‘short.’ Panoan Culino tukuča ‘short.’ Problems: symbolic? Semantics?: small / short.
57 SON Charruan: Chana, Guenoa ineu. Guaicuru[an]: Guachi [Guachi is not a Guaicuruan language] inna. Vilela ina-hmi (Pelleschi [source]), ina-ke ‘son, daughter’ (Gilij [source]), hina-kis (Fontana [source]). Problems: Classfication of Guachi.
58 SOUR Mataco[an]: Choroti paši
59 SWIM Guaicuru[an]: Pilaga ubogai. Moseten
60 THIN Lule kam. Moseten kum. Cf. Mataco[an]: Vejoz [Wichí] čemsa- ‘small.’ Problems: semantics: thin / small.
61 URINE [Mascoyan:] Lengua jis(-weji) ‘urinate.’ Lule
62 WEAK Lengua [Mascoyan]: Mascoy jil, jel-k. Mataco[an]: Mataco [Wichí] jel ‘weak, tired’ [y- ‘3rd Person Pronoun’, root -eɬ ‘to tire’]. Problems: morphological analysis: failure to segment off Nivaclé y- makes the forms appear similar when there is little similarity. Short forms.
139
What evidence (if any) do you find that might support a possible genetic relationship among the languages compared here? There are problems with some forms in nearly all of the proposed cognate sets. In many cases so many forms in the languages compared have problems that in effect the entire set must be eliminated. When one tries to establish sound correspondences across the languages that are proposed as belonging to the grouping, there are insufficient remaining forms to be able to establish any reliably recurring sound correspondences among Panoan, Tacanan (Pano- Tacanan), Mosetén, Matacoan, Guaicuruan, Charruan, Lule, Vilela (Lule-Vilela), and Mascoyan (with the isolates Payagua and Guachi slipped in). For example, Charruan forms (not even all of whihc are from the same Charruan language) appear in only four lexical sets (5, 16, 40, 57), far to few to establish sound correspondences for Charrua. The same story holds for Vilela, represented in only five lexical sets (16, 39, 44, 46, 57), and some of these are questionable. For the other languages, too few forms remain that are not seriously challenged so it is not possible to build a case for genetic relationship based on them.
140
EXERCISES FOR CHAPTER 16 LINGUISTICS PREHISTORY
Exercise 16.1 Proto-Chibchan cultural inventory
Chibchan is a language family of more than twenty languages spoken in southern Central America and northern South America. What do the following reconstructed Proto-Chibchan vocabulary items reveal about the culture of Proto-Chibchan speakers?
1. *dihke ‘to sow’ 2. *ta1 ‘cultivated clearing’ 3. *ike ‘manioc’ 4. *tuʔ ‘tuber, yam’ 5. *apì ‘pumpkin, squash’ 6. *e, *ebe ‘maize’ 7. *du, *dua1 ‘tobacco’ 8. *tã1 ‘gourd rattle’ 9. *toka ‘gourd cup’ (Data from Constenla Umaña 2012.)
Proto-Chibchan speakers were agriculturalists, apparently cultivating several plants, including maize, squash, gourds, and manioc.
Exercise 16.2 Proto-Tupían cultural inventory
Tupían is a large language family of South America with c.70 languages, spoken in Argentina, Bolivia, Brazil, Colombia, French Guiana, Paraguay, Peru, and Venezuela. (Tupí-Guaranían is a large subgroup of this family.) Examine the following reconstructed Proto-Tupían vocabulary items; what are you able to say about the culture of the speakers of Proto-Tupían based on these?
1. *awai ‘yam (Dioscorea sp.)’ 2. *čekw ‘to pound’ 3. *čɨt ‘to bake’ 4. *čʔam ‘rope’ 5. *čét ‘digging stick’ 6. *ekw ‘house’ 7. *ekwat ‘village patio’. 8. *ekwen ‘door’ 9. *ekwʔɨp ‘arrow’ 10. *ẽɾi ‘hammock’ 11. *eʔe ‘to grate’ 12. *ɨɾju ‘basket’
141 13. *ɨʔa ‘calabash (gourd)’. 14. *kuɾua ‘pumpkin’ 15. *mani ‘manioc’ 16. *mõj ‘to cook’ 17. *ŋo/ŋe ‘cultivated field’ 18. *pe ‘tobacco’ 19. *ɾjuku ‘achiote (Bixa orellana)’ 20. *t’ap ‘thatch’ 21. *upap ‘bed, lying place’ 22. *wamu/wamuã ‘shaman’ 23. *waʔẽ ‘ceramic pot’ 24. *wekẽʔa ‘fish trap’ 25. *wetjɨ́k ‘sweet potato’ 26. *wɨ ‘ax’ 27. *wɨp ‘to cook, bake’
Proto-Tupían speaker’s cultural inventory appears to have included: agriculture (yam, gourd, pumpkin, manioc, sweet potato, cultivated field, digging stick), ceramics (ceramic pot), fishing (fish trap), an assortment of implements (rope, arrow, hammock, basket, thatch, bed), and shamanism (shaman).
Exercise 16.3 Proto-Muskogean environment
Languages of the Muskogean family include Choctaw, Chicasaw, Creek, Mikasuki, and others, once spoken widely in the southeastern US. What can you say about the Proto-Muskogean speakers’ knowledge of their environment and possibly about the Proto-Muskogean homeland based on the fact that terms for the following can be reconstructed to Proto-Muskogean? (Only the glosses of the reconstructed items are given.
1. apple (crab-apple or persimmon) 2. bat 3. bee 4. beetle 5. bluejay 6. briar 7. briar (blackberry) 8. buckeye 9. buffalo 10. chestnut 11. chicken snake 12. chickenhawk 13. chigger 14. chipmunk 15. clam /spoon 16. copperhead
142 17. corn (maize) 18. cotton 19. crane (whooping crane) 20. crawfish 21. cricket 22. deer 23. dove 24. duck 25. falcon 26. flea 27. frog 28. goose 29. grape 30. grasshopper 31. hackberry 32. haw 33. heron 34. hickory 35. hoe /plough 36. honey locust 37. hoot owl 38. horned owl 39. hornet /wasp 40. horsefly 41. hummingbird 42. katydid 43. lamb's quarters (chenopodium) 44. leech 45. lightning bug 46. lizard 47. locust/cicada 48. louse 49. martin 50. milkweed 51. mole 52. moss 53. mountain lion 54. muddauber 55. mulberry 56. muscadine grape 57. mushroom 58. oak (3 kinds) 59. onion 60. opossum 61. palmetto 62. pear
143 63. perch 64. pigeon 65. pokeweed 66. potato 67. prickly 68. pumpkin 69. quail 70. rabbit 71. redbud 72. redheaded woodpecker 73. screech owl 74. skunk 75. slippery elm 76. snake 77. spider 78. squirrel, 79. stinging plant (poison ivy?) 80. tadpole 81. thrush 82. trout 83. turtle 84. turtle (soft-shelled) 85. walnut 86. water lily 87. whippoorwill 88. wildcat, 89. woodpecker 90. worm 91. wren 92. yellowhammer (From Broadwell 1992.)
The homeland of Proto-Muskogean speakers can be determined by examining the geographical distribution of the species in this list to see where they intersect. Broadly distributed species, such as ‘bat’, ‘bee’, ‘flea’, ‘frog’, do little to localize the homeland. The ranges of some of these which are more helpful for localizing the homeland are: buckeye (Aesculus glabra): ranges south to eastern Kansas, southwestern Oklahoma, central Texas; east to western Arkansas, Tennessee, and central Alabama, one location in eastern Mississippi. buffalo (Bison bison): Their range was roughly a triangle between the Great Bear Lake in Canada's far northwestern Canada, south to the Mexican states of Durango and Nuevo León, and east along the western boundary of the Appalachian Mountains. chestnut (Castanea dentata): ranged from Maine and southern Ontario to Mississippi, and from the Atlantic coast to the Appalachian Mountains and the Ohio Valley. chicken snake (Pituophis melanoleucus melanoleucus): its range extends from southern North
144 Carolina west through South Carolina to northern Georgia, eastern Tennessee, southeastern Kentucky and south into Alabama. Disjunct populations exist in southern New Jersey, west-central Virginia, adjacent West Virginia, central Kentucky and southwestern Tennessee. This subspecies intergrades with Pituophis melanoleucus mugitus in South Carolina, Georgia, and Alabama. copperhead (Agkistrodon contortrix): Alabama, Arkansas, Connecticut, Delaware, Florida, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Texas, Ohio, Oklahoma, Maryland, Massachusetts, Mississippi, Missouri, New Jersey, New York, North Carolina, Pennsylvania, South Carolina, Tennessee, Virginia and West Virginia. In Mexico it occurs in Chihuahua and Coahuila. crab-apple, persimmon (Diospyros virginiana): it ranges from New England to Florida, and west to Texas, Louisiana, Oklahoma, and Kansas. honey locust (Gleditsia triacanthos): a deciduous tree native to central North America, ranging from southeastern South Dakota to New Orleans and central Texas, and as far east as eastern Massachusetts. mulberry (Morus rubra): native to eastern North America, from Ontario and Vermont south to southern Florida and west to southeast South Dakota and central Texas. muscadine grape (Vitis rotundifolia): native to the southeastern US; its natural range was: Alabama, Arkansas, the Carolinas, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Missouri, Mississippi, Oklahoma, Tennessee, Texas, Virginia, and West Virginia. palmetto (Sabal, a genus of New World palms with many species), native to the warm temperate to tropical regions of the Americas, from southeastern US through the Caribbean, Mexico and Central America to Colombia and Venezuela in northern South America. pokeweed (Phytolacca americana): native to eastern North America, the Midwest, and the Gulf Coast, with more scattered populations in the far West.
The ranges of these species all overlap in the southeastern US perhaps including also the Gulf Coast.
Exercise 16.4 Proto-Uto-Aztecan cultural inventory
Uto-Aztecan is a large language family with languages stretching from Oregon to Nicaragua. Examine the following reconstructed Proto-Uto-Aztecan terms. What can you say about the culture of the speakers of Proto-Uto-Aztecan? What might you be able to say about the probable homeland of Proto-Uto-Aztecan?
1. *aCta ‘bow, atlatl’ 2. *amu ‘agave (yucca plant)’ 3. *amu ‘hunt’ 4. *ayaw ‘gourd, squash’ 5. *ayo ‘turtle’ 6. *hulapɨ ‘badger’ 7. *huma ‘flour, meal’
145 8. *hutsa ‘arrow’ 9. *kakV ‘crow’ 10. *kɨmal ‘blanket’ 11. *koloka ‘beads, necklace’ 12. *kuma ‘tool (for poking, cutting), knife’ 13. *kuna ‘sack, bag’ 14. *kusa ‘bag, sack’ 15. *kwika ‘sing’ 16. *kwisa ‘eagle’ 17. *maCta ‘grinding stone (for seeds, grains), mortar’ 18. *motoʔo ‘squirrel’ 19. *muCta ‘cholla cactus’ 20. *naka ‘mountain sheep (meat)’ 21. *osa ‘paint, draw’ 22. *pakaN ‘read’ 23. *paʔtsi ‘seed’ 24. *piŋa ‘grind’ 25. *pipa ‘tobacco’ 26. *pɨtsɨN ‘duck’ 27. *piʔa ‘gather, pick’ 28. *pona ‘play music, play drum’ 29. *pota ‘cottonwood tree’ 30. *putsi ‘seed, pit’ 31. *sayo ‘enemy’ 32. *sikuli ‘peyote’ 33. *taka ‘fruit’ 34. *tapi ‘hawk’ 35. *tɨkpa ‘cutting tool, obsidian, knife’ 36. *tɨsoli ‘quail’ 37. *topi ‘cottontail rabbit’ 38. *toptu ‘dance’ 39. *tsal ‘loincloth’ 40. *tsik ‘basket’ 41. *tsɨka ‘duck’ 42. *tuʔi ‘grind, flour’ 43. *tuʔtsa ‘hummingbird’ 44. *wa(s)sa ‘crane’ 45. *wi-talo ‘roadrunner’ 46. *wika ‘digging stick’ 47. *wiki ‘rope, string’ 48. *wipula ‘belt, sash’ 49. *wokoN ‘pine, pine tree’ 50. *yawa ‘dance’ (From Brian Stubbs 2011.)
146 Proto-Uto-Aztecan culture: Hunting-and-gathering subsistence: ‘bow, atlatl’ ‘hunt’ ‘arrow’ ‘turtle’ ‘mountain sheep (meat)’ ‘seed’ ‘duck’ [2 species] ‘gather, pick’ ‘seed, pit’ ‘fruit’ ‘cottontail rabbit’ The presence of ‘grinding stone, mortor’, ‘grind’, ‘flour, meal’, ‘digging stick’, and ‘gourd, squash’ might at first seem to suggest the possibility of agriculture; however, hunter- gatherers also collect wild seeds, grind them, grind them on grinding stones in some instances, make meal or flour from them, and often use digging sticks, and wild gourds were known (‘squash’ appears to be a later semantic shift from something that earlier meant ‘wild gourd’ or something similar). So this is not good evidence that Proto-Uto-Aztecan speakers had agriculture.
Material culture: implements and clothing: ‘bow, atlatl’ ‘arrow’ ‘tool (for poking, cutting), knife’ ‘cutting tool, obsidian, knife’
‘grinding stone, mortar’ ‘digging stick’
‘sack, bag’ [1] ‘bag, sack’ [2] ‘basket’ ‘rope, string’
Clothing, adornment and ritual: ‘beads, necklace’ ‘blanket’ ‘loincloth’ ‘belt, sash’
‘paint, draw’ ‘tobacco’ ‘play music, play drum’ ‘dance’ Tobacco, music, and dance were all associated with ritual.
147
The cultural inventory from the reconstructed lexical items is consistent with non- agriculturalist hunting-and-gathering culture, with no evidence of metal or ceramics (though one cannot argue for what the culture may have had or lacked from the absence of evidence, only from the presence of evidence).
Proto-Uto-Aztecan homeland: within the area where the ranges of the following intersect: ‘agave (yucca plant)’ (genus of monocots, family Asparagaceae, subfamily, with about 208 species: chiefly in Mexico, also native to the southern and western US, and central and tropical South America. ‘badger’ (American badger Taxidea taxus): found in the western and central US, northern Mexico, central Canada, and areas of southwestern British Columbia. ‘cholla cactus’ (Cylindropuntia is a genus of cacti, family Cactaceae), about 35 species: native to the southwest and south central United States, Mexico, and the West Indies. ‘mountain sheep’ (Ovis canadensis): through western North America as far south as Baja California and northwestern mainland Mexico. There are several subspecies. Sierra Nevada bighorn sheep (Ovis canadensis sierrae) is a genetically distinct subspecies that only occurs in the Sierra Nevada, California. Desert Bighorn Sheep (Ovis canadensis nelsoni) occurs throughout the southwestern desert regions of US and Mexico. ‘cottonwood tree’(probably Populus fremontii ‘Fremont cottonwood’): occurs in California east to Utah and Arizona and south into northwest Mexico, similar to the eastern cottonwood (Populus deltoides), which occurs throughout the eastern and in extreme southern Canada. ‘peyote’ (Lophophora williamsii): native to southwestern Texas and Mexico, primarily in the Chihuahuan desert and in the states of Tamaulipas and San Luis Potosí. ‘roadrunner’ (probably Greater Roadrunner, Geococcyx californianus): can be found in California, Arizona, New Mexico, Texas, Nevada, Utah, Colorado, Oklahoma, and rarely in Kansas, Louisiana, Arkansas and Missouri, and in the Mexican states of Baja California, Baja California Sur, Sonora, Sinaloa, Chihuahua, Durango, Jalisco, Coahuila, Zacatecas, Aguas Calientes, Guanajuato, Michoacán, Querétaro, México, Puebla, Nuevo León, Tamaulipas, and San Luis Potosí.
These species place the Proto-Uto-Aztecan homeland in a region where the ranges of these species overlap – all are found in the US southwest and northern Mexico, though some have a more northerly distribution and some more southerly ranges. Therefore the Uto-Aztecan homeland must have been in the US southwest-northern Mexico region.
148