GRS LX 700 Morphology Language Acquisition  In L1A, we observe that kids don’t always and provide all of the morphology that adults do. Linguistic Theory  Traditionally, it was assumed that kids are learning the morphology and the Week 13. L2 morphology v. and that at some point they got it (say, functional projections when they provide correct morphology 90% of the time when it was required).

Morphology German and L1A CP  A major recent development in the study of how DP C′  So, in German. kids come to know the (by now, known to be  When a 2-year-old fabulously complicated, but yet relatively C+I IP John ate uses a finite , it language-independent) system of syntax was in — I′ goes in second the observation that morphological errors are by VP — position; when a 2- no means random. year-old uses a  In particular, in a large number of languages, V′ nonfinite verb it remains at the end of what seems to happen is that kids produce DP — nonfinite forms of the verb—but along with that the sentence (after the ). comes the syntax associated with non-finiteness. lunch

Functional categories Functional categories

 So, even though kids will sometimes use  Rephrasing a bit, what we’re talking about is essentially nonfinite , they know the difference between the structural complexity of the learner’s (L1A/L2A) finite and nonfinite verb and know how the knowledge (at a given point). grammar treats each kind. They are using T correctly. They just sometimes pick the wrong  It has been pretty well established by theoretical (nonfinite) one. linguistics that adult native languages are quite complex,  Now, adult L2’ers also drop a lot of morphology, containing functional like AgrP, TP and CP, and will produce nonfinite forms… there is a lot of support for this idea that most if not all parametric differences stem from properties of the  This raises the question (in the general ballpark abstract functional morphemes (often reflected in surface of “how much is L2A like L1A?”) as to whether morphology). second language learners show this effect as well.

1 Functional categories Functional categories

 Verb movement (if it conforms to the rules of adult native-speaker verb movement, anyway)  The question we’re about to look at is whether adult serves as evidence for this complex functional second language learners also have this same complex structure, since the verb moves into a functional structural knowledge in their IL. Do L2’ers “know about TP” in other words? head (T, for example).  The evidence we just reviewed suggests very  Note that if L2’ers can usually produce sentences which strongly that kids learning German and French are grammatical in the TL but yet don’t “follow the produce sentences which comply with the rules rules” which are associated with that structure (i.e. that of adult syntax (that make reference to this only finite verbs move to T), we do not have evidence complex functional structure). Kids seem to that their mental representation of these sentences “know about” the TP and the CP and the rules includes the higher functional phrases like TP. that pertain thereto.

Prévost and White (1999, 2000) Prévost and White  Prévost and White try to differentiate two  Prévost and White (1999, 2000) investigated this possibilities of what their data might show, given very question, and here’s what they found. that second language learners sometimes use inflected verbs and sometimes don’t.  Impairment Hypothesis. The learners don’t really  Like kids do during L1A, second language learners (consistently) understand the or how to use it. will sometimes omit, and sometimes provide, Their knowledge of inflection is “impaired”. Their trees inflection (tense, agreement) on the verb. don’t contain the functional XPs.  Missing Surface Inflection Hypothesis. The learners will sometimes pronounce finite verbs in their form  However, it is different from L1A in that lack of finite (the verbs act finite, the functional XP’s are there, but the inflection on the verb does not seem to correlate with learner couldn’t find the right inflected form in his/her being treated syntactically as an infinitive. lexicon in time, so s/he used the nonfinite form). The nonfinite form is essentially a default.

Prévost and White Prévost and White

 Possibility 1 (impairment) suggests basically no correlation between verb movement and  P&W looked at spontaneous speech data inflection. from two adults learning L2 French (from Moroccan Arabic, after a year) and two  Possibility 2 (mispronouncing a adults learning L2 German (from Spanish by using its nonfinite form) predicts that and Portuguese, after 3 months). Monthly  When the finite form is pronounced, the verb interviews followed for about 2 years. will definitely be (and act) finite—it will move.  When the nonfinite form is pronounced, it might act finite or nonfinite.

2 Prévost and White found… Prévost and White  Almost no finite (inflected) verb forms in non-finite contexts.  When verbs are marked with inflection, they systematically (overwhelmingly) appear before negation (i.e., they move).  P&W’s data supports the hypotheses that:  Many of nonfinite forms used in finite contexts (used finitely,  (These) second language learners know the difference moved). between finite and nonfinite verbs.  They know that finite verbs move, and that nonfinite verbs Oblig. Fin Oblig. Nonfin do not move. +Fin -Fin -Fin +Fin  The only real errors they make are essentially lexical retrieval errors (errors of pronunciation), pronouncing verbs A(F) 767 243 278 17 which are abstractly finite in their infinitive form. Z(F) 755 224 156 2  One question: Why the infinitive? Is it really an unmarked A(G) 389 45 76 7 form universally? Does it depend on what the citation form is? Is it due to the language-particular morphology. Z(G) 434 85 98 6

L2A and L1A L2A and L1

 One thing this tells us is that, despite  We don’t know really to what extent “UG” possible appearances to the contrary, played a role, based only on this—after all, second language learners’ interlanguages we know that the L1 had the full structural are quite systematic and complex, and the complexity of a natural language, including L2 learners have the same kind of abstract the distinction (perhaps abstract) between structural knowledge incorporated into finite and nonfinite, and including (perhaps their IL that we can argue for in the case of abstract) subject agreement, etc. There’s no reason that knowledge of the distinction L1 learners. between finite and nonfinite couldn’t simply carry over (“transfer”) to the IL during L2A.

Morphology ≠ syntax Morphology ≠ syntax

 This suggests that morphology is rather distinct  Various other studies describe a similar from syntax. It is possible to have the syntax dissociation; obligatory subjects, subject right and the morphology wrong. And to some case, and verb position are all governed by extent, morphology is not provided by UG, must syntactic features/parameters attributed be learned, and moreover must be retrieved. to functional projections. And while L2’ers  The view of Distributed Morphology under which morphology is a separate system given seem to get these right, they are the task of pronouncing a syntactic structure inconsistent with the morphology. (See (and which allows for the sort of defaults we White ch. 6; Lardière, White, Schwartz, seem to see) seems well suited to describe this. Prévost, …)

3 Schwartz (2002) Schwartz (2002)

 Last year at the BUCLD, Bonnie Schwartz  Schwartz concluded that presented data of this sort looking at the gender agreement and definiteness properties of Dutch  child L2 is like child L1 wrt morphology DPs, with the aim being to determine whether  child L2 is like adult L2 wrt syntax child L2 acquisition was more like child L1  Again, a dissociation between acquisition or more like adult L2 acquisition. morphology and syntax.  What she found was that in terms of overgeneralizing morphology (overuse of uninflected ), adult L2’ers did it, but  Why? Morphology is surface-evident and neither child L1’ers nor child L2’er did. But in terms of word order, both kinds of L2’er went frequent, why is there such difficulty? through a word order stage not attested in child L1’ers’ development.

thoughts re: Schwartz (2002) thoughts re: Schwartz (2002)

 Jeff Lidz brought up the question of  Harald Clahsen brought up an interesting point whether this might be due not so much to with respect to processing: there are processing results that indicate that adult L2’ers “need morphology, but to a phonological effect. longer” to process incoming data. While I’m not Either in terms of an input filter (like the sure exactly what studies he had in mind, taking French discussion earlier) or in terms of a that as given, perhaps the problem with morphology is that it just “comes too fast.” In production constraint. Phonological the same kind of way that phonological filters problems could in many ways mimic might keep morphological marking out of the morphological problems. “input data”, processing constraints might also have this effect.

 Language attrition

   It is a very common phenomenon that, having learned an L2 and having become  quite proficient, one will still “forget” how   to use it after a period of non-use.    While very common, it’s not very  surprising—it’s like calculus. If L2 is a skill like calculus, we’d expect this. 

4 L1 attrition UG in L2A

 Much more surprising is the fact that sometimes under the influence of a dominant L2, skill in the L1  We’ve looked at the questions concerning seems to go. whether when learning a second language, one can adapt the “parameter settings” in the new knowledge to the target settings (where they  Consider the UG/parameter model; a kid’s LAD faced with PLD, automatically sets the parameters differ from the L1 settings), but this is even more in his/her head to match those exhibited by the dramatic—it would seem to actually be altering linguistic input. L1 is effortless, fast, uniformly the L1 settings. successful… biologically driven, not learning in the  It behooves us to look carefullier at this; do normal sense of learning a skill. attrited speakers seem to have changed  So how could it suffer attrition? What are you left parameter settings? with?

Italian→English Italian, null subjects

 Q: Perchè Maria è uscite?  Italian is a “null subject” language that allows the subject to be dropped in most cases where in ‘Why did M leave?’ English we’d use a  A1: Lei ha deciso di fare una passeggiata.  (Possible to use a pronoun in Italian, but it conveys  A2: Ha deciso di fare une passenggiata. something pragmatic: contrastive focus or change in topic) ‘She decided to take a walk.’  English is a “non-null-subject” language that does not allow the subject to be dropped out,  Monolingual Italian speaker would say A2, are required (even sometimes “meaningless” like it or there). Not required that but English-immersed native Italian a pronoun signal a change in topic. speaker will optionally produce (and accept) A1. (Sorace 2000)

Reverse errors unattested Postverbal subjects

 Q: Perchè Maria è uscite?  Q: Chi ha starnutito? ‘Who sneezed?’ ‘Why did Maria leave?’  A1: Gianni ha starnutito.  A: *Perchè Ø è venuto a prederla.  A2: Ha starnutito Gianni. ‘Because (Gianni) came to pick her up.’

 Native speakers would say A2 due to the  That is, they don’t forget how to use null narrow focus; attrited speakers will subjects so much as they broaden the produce/allow A1 as well. contexts in which they can use overt pronouns.

5 L1 attrition L1 attrition

 It seems that the acceptability of overt pronouns  Same goes for postverbal subjects—it is a (in the L1 “attriters”) broadens compared to marked option for languages, and the L1 their L1, the acceptability of null pronouns becomes more restricted. seems to be retreating to the unmarked.  Pronouns in a null subject language are marked—they are restricted to particular discourse contexts ([+topic shift], according to  Like with pronouns, it seems to be not a Sorace). question of grammaticality but a question  What seems to happen is that the pronouns of felicity. revert to the unmarked case ([±topic shift] like in English).

L1 attrition L1 attrition

 Certain areas of the L1 grammar are more  So, we’re left with a not-entirely- susceptible to this kind of attrition then others. inconsistent view of the world.  Parameter settings in L1 appear to be safe,  Sorace notes that the observed cases of attrition of this sort seem to be the ones involved with but the discourse-pragmatic constraints discourse and pragmatics, not with fundamental seem to be somehow susceptible to high grammatical settings. (The attrited Italian is still exposure to conflicting constraints in other a null-subject language, for example—null subjects are still possible and used only in places languages. where null subjects should be allowed).

Language mixing  (Spanish-English)    No, yo sí brincaba en el trampoline when I was a senior.  ‘No, I did jump on the trampoline when I was a senior.’

   La consulta era eight dollars. ‘The office visit was eight dollars.’  

 Well, I keep starting some. Como por un mes todos los días  escribo y ya dejo.  ‘Well, I keep starting some. For about a month I write everything and then I stop.’

6 But it isn’t random… Prior efforts

 *El viejo man The old man  Several proposals have been offered to account  *The old hombre El hombre viejo for what are good mixes and what aren’t, but it appears to be a hard problem. Very famous  *The viejo hombre attempt by Poplack (1980, 1981):  The equivalence constraint. Codes will tend to  *She sees lo. be switched at points where the surface structure of the languages map onto each other.  Certain mixes are not considered to be possible  The free morpheme constraint. A switch may by fluent bilinguals. occur at any point in the discourse at which it is  How can we characterize what mixes are possible to make a surface constituent cut and possible vs. impossible? still retain a free morpheme.

Poplack Problems for Poplack

 Looking at the constraints on code-switching of this sorts can help us understand the nature of  Equivalence and Free Morpheme Constraints: (at least fluent) bilingual language Accounts for *estoy eatiendo, but leaves representation. unexplained:  The students habian visto la pelicula italien.  One odd thing about Poplack’s constraints is  *The student had visto la pelicua italien. that it implies that part of UG is dedicated to  *Los estudiantes habian seen the Italian movie. mixing. The Free Morpheme Constraint and Equivalence Constraint are only constraints on mixing two grammars. Is UG built specifically  Motrataroa de nin kirescataroa n Pocajontas for bilinguals? Ref-treat-vsf about this 3s-3os-rescue-vsf in P. ‘It deals with the one who rescues P.’

Problems for Poplack? Problems for Poplack

 *Tú tikoas tlakemetl  *El no wants to go 2sg 2s-3Os-buy-fut garment-pl-nsf  *He doesn’t quiere ir. (‘You will buy clothes’)

 *No nitekititoc not 1s-work-dur (‘I’m not working’)  El kikoas tlakmetl he 3S-3Os-buy-fut garment-pl-nsf ‘He will buy clothes’  Amo estoy trabajando not be.3s work-dur ‘I’m not working’

7 MacSwan 1999 MacSwan 1999

 Perhaps the most currently comprehensive  The broad (“minimalist”) approach to and promising account, building on recent grammar takes language to consist of two developments in syntactic theory. primary components.

 One of the basic premises is that language  Computational system (builds trees), parameters are properties of lexical items language invariant. (not of a language-wide grammar). E.g., verb-movement is due to a property of the  Lexicon, language particular. Functional tense morpheme in French, not shared by elements of the lexicon encode the the tense morpheme in English. parameters of variation.

MacSwan 1999 MacSwan 1999

 MacSwan’s proposal is that there are no constraints on code mixing over and above constraints found on  The model of code mixing is then just like monolingual sentences. monolingual speech—the only difference  (His only constraint which obliquely refers to code mixing is the being that the words and functional one we turn to next, roughly that within a word, the language must be coherent.) elements are not always drawn from the lexicon belonging to a single language.

 We can determine what are possible mixes by looking at the properties of the (functional elements) of the lexicons  Where requirements conflict between of the two mixed languages. languages is where mixing will be prohibited.

Clitics, bound morphemes Clitics, bound morphemes

 Some lexical items in some languages are  Clitics generally cannot be stressed. clitics, they depend (usually  *John’S book phonologically) on neighboring words.  *I couldN’T go. Similar to the concept of bound morpheme.  John’s book.  Clitics generally form an inseparable unit  I shouldn’t go. with their host.  Shouldn’t I go?  Clitics essentially fuse with their host.  Should I not go?  *Should I n’t go?

8 Spanish no Nahuatl amo

  It turns out that Spanish no appears to be a In Nahuatl, amo ‘not’ does not appear to clitic (despite spelling conventions). be a clitic.  ¿Qué no dijo Juan? ‘What didn’t J say?’   *¿Qué sólo leyó Juan? (‘What did J only Amo nio amo niktati nowelti. read?’) Not 1s-go not 1s-3Os-see my-sister ‘I’m not going to not see my sister.’  *¿Qué meramente leyó Juan? (‘What did J merely read?’)  *Juan no ha no hecho la tarea. (‘J hasn’t not done the task.’)

Spanish-Nahuatl mixing MacSwan 1999

 *No nitekititoc  MacSwan proposes essentially that it is not possible to not 1s-work-dur (‘I’m not working’) code-mix within a (word-like) phonological unit. Essentially a  Amo estoy trabajando restriction on what are “pronouncable” trees.  Idea: phonology operates as a set of ordered rules which are not be.3s work-dur ‘I’m not working’ ordered differently in different languages—you can’t run both sets of rules at once, hence the result if you tried would be unpronounceable.  Now, we can begin to make sense of the  Since Spanish no fuses with the following verb, it can’t difference in possible mixes at the point of be followed by a Nahuatl verb. negation between Spanish and Nahuatl.  Since Nahuatl amo does not fuse with the following verb, it is free to be followed by a Spanish verb.

English-Spanish Agreement

 This also explains Spanish-English (well,  In languages that code agreement between subject and Spanish-anything) verb, it also appears that mixing is only possible where the agreement relationship is not disrupted.  *El no wants to go

 *He doesn’t quiere ir.  What about English-Spanish?  English negation: agreement appears on do.  *He doesn’t quiere ir.  Spanish negation: agreement appears on the verb.

 *He doesn’t wants to go.  You can’t have extra agreement: one subject, one agreement. They need to match.

9 Agreement Agreement

 *Yo nikoas tlakemetl  Ni-k-koa-s ‘I will buy’ I 1s-3Os-buy-fut garment-pl-nsf  Ti-k-koa-s ‘You will buy’ (‘I will buy clothes’)  Ø-k(i)-koa-s ‘He/she wlll buy’  *Tú tikoas tlakemetl you 2s-3Os-buy-fut garment-pl-nsf (‘You will buy clothes’)  Also relevant: Spanish marks and agrees with gender but Nahuatl does not distinguish masculine from feminine.  Él/Ella kikoas tlakemetl He/She 3s-3Os-buy-fut garment-pl-nsf  Spanish pronouns have gender specification. The Nahuatl ‘He/She will buy clothes’ verb does not. They can only be compatible (match) if there is no Nahuatl agreement morpheme.

Spanish-Catalan-Greek Spanish-Catalan-Greek

 Spanish and Catalan both have two  Yo vull mengar el dinar (S-C) genders, masculine and feminine.  Jo queiro comer la cena (C-S)  Greek has three genders, masculine,  *Ego vull mengar el dinar (G-C) feminine, neuter.  *Ego queiro comer la cena (G-S)

 Predicts: Mixing subjects and verbs  … between the three languages is only possible between the gender-compatible languages.

Mixing and L2A? 

 Code mixing as discussed so far is generally a   property of the speech of fluent bilinguals (often native bilinguals) and reflects properties of universal  language knowledge.      We can now return to our old question and ask: Does the knowledge of second language learners also have  the restrictions on code mixing? To the extent that this is “part of UG”, is this aspect of UG active for L2’ers?  For the future—I’m not aware of studies on L2A.

10 Some major views on L1A: Case errors L1A/syntax  Kids will sometimes make case errors  Radford/Guilfoyle/Noonan: kids lack with the subject (until around 2). functional elements initially, have only  Me got bean. lexical elements.  In English, accusative (me) is the “default.”  Wexler: kids have access to all the same  Very often taken to indicate a subject not in grammatical elements that adults do. SpecIP (a.k.a. SpecAgrSP). No IP? (Radford)  Rizzi: kids have “truncated trees” Sometimes IP and above (Rizzi, Vainikka)? No AgrSP? (Wexler)  Vainikka: kids “grow trees”

L1A: Null subjects L1A: Optional

 Kids will also often drop out subjects, even in  In many languages, kids will allow languages where null subjects are not allowed. nonfinite verbs in root sometimes,  Hyams (1986): Mis-set parameter; they’re speaking early on (up to a little after 2). Italian initially.  Kids who are learning null subject languages drop  NS/OI? Wexler (1998) suggests that more subjects than kids who are learning non-null there’s a strong correlation between lack of subject languages. OI’s in 2-year-old speech and being a null  Bloom: Long sentences are harder, drop what you can. The beginning of a sentence is more susceptible. subject language.  Wexler/Hyams: Kids drop more subjects with  True? Or are OI’s just extra-rare in null subject nonfinite verbs. PRO. Sometimes topic drop with languages (correlation with more elaborate finite verbs, where “topic” isn’t yet grasped. inflection?).

L1A: Finite vs. nonfinite Some stories about OIs

 During Optional Infinitive stage, kids with  Rizzi: until maturation of Root=CP, trees OI’s treat finite verbs like finite verbs and truncated sometimes below tense. nonfinite verbs like nonfinite verbs.  Wexler/Schütze: Syntax intact, but something prohibits the same (subject) DP from licensing  German (Poeppel & Wexler): V2 for finite both TP (finite tense) and AgrP (Nom case). verbs, final V for nonfinite verbs.  Radford: Kids don’t use functional categories at this point (yet, leaves the “finite verbs act finite”  French (Pierce): Verb before pas for finite data unexplained). verbs, verb after pas for nonfinite verbs.  Legendre et al: Kids minimize the number of functional projections, basically same outcome as Schütze & Wexler.

11 L1A: Principles B and P L1A: A-chains, passives

 Even older kids seem to allow co-reference  Kids are also purportedly slow to master in apparent violation of Principle B: Mary passives and unaccusatives. saw her.  Borer & Wexler (1987): This is maturation of the  Chien & Wexler, then Thornton & Wexler, ability to represent “A-chains”—more show that when quantifier binding is specifically, the ability to move an object-type available (and thus requires coindexation), thing into a subject-type position (non-local Principle B is respected. assignment of θ-roles).  Principle P is slow in coming (matures?),  Babyonyshev et al. (1998) show kids have which says coreference --> coindexation. trouble with the genitive of negation.

L1A: A-chains etc. L1A: Negation outside of IP

 Some possible reasons for skepticism on  Kids for a while seem to have trouble with this: negation outside the IP, and repair their  Snyder, Hyams, Crisma (1994): French kids utterances so that it remains inside get auxiliary selection right with reflexive (usually in an adult-ungrammatical way). clitics:  What kind of bread do you don’t like?  Le chienj si’est [ ti mordu tj ].  VP-internal subjects  Where he couldn’t eat the raisins?  Korean negation misplacement seems to differentiate unergative/transitive from unaccusatives. (not previously discussed)

L1A: Syntax L2A: What can we say?

 In general, the errors kids are making seem to be very systematic.  Certain things are required to explain L1A.  They seem to know many aspects of the  Kids don’t get negative evidence  or if they do, it is inconsistent, it is noisy, and moreover grammatical system, allowing us to pinpoint (if sometimes when we try to give them negative evidence, they we look closely enough and ask the right ignore it. questions) what parts don’t seem to be working.  The kids must be able to learn a system that assign * to  A-chains (or “dethematization of an external arg.”). some sentences, based only on positive evidence.  Using a [D] feature twice to check functional features.  Conclusion: Universal Grammar constrains the kinds of languages there can be, those languages cannot generate  Allowing negation in C. certain kinds of sentences (hence: *).  Requiring coreference to imply coindexation.

12 L2A: What can we say? L2A: What can we say?

 L2A is generally much harder, more  L1A: Languages differ from one another. conscious, slower, less successful.  Something needs to be learned from the environment.   Yet much of the grammatical system seems common What’s different about L2A? Did UG across languages. disappear? Did the LAD disappear?  Languages can be thought of as varying not in the system  Question: What is the state of the L2’ers (the principles) but in the parameters. knowledge about the L2?  The kids, who learn their native language so fast, must have some help setting the parameters. A Language  Does this conform to what UG would Acquisition Device (LAD) designed to choose among the allow? options made available by UG.

L2A: UG-accessibility L2A: Transfer

 If the IL is UG-constrained, what is the initial  In general, it seems that the evidence starting assumption? points to the interlanguages being  Is it some kind of general default setting for all the allowable human languages. This could parameters (likely to be a “subset” grammar from either be influence from UG which all other grammars can be learned via (constraining possible languages) or position evidence alone)? because the IL is a variation on L1.  Is it just carrying over the parameter settings from  Can we tell? Look at parameter settings: L1? Does IL represent a different option  Some combination of these? from L1?

L2A: Interlanguage = L2A: Tricks L1+prescriptive rules?  Is the IL just L1 plus some prescriptive rules  In order to look properly at parameters, we need (LLK)? (Fundamental Difference) to know what they are. And what a “default”  Or does the IL actually show resetting of setting might be. This turns out to be hard. parameters?  Pro-drop parameter. Default: Drop subjects?  Resetting should entail: cluster of properties Subset learnable? Correlated with anything else? comes with new value (again requires that we  Binding Theory Governing Category? Default? know what the parameters, values, clusters are) Language-wide? Strictly predictable from  If we can find a non-L1, non-L2, but UG- morphology? available option in the IL, that also suggests parameter setting.

13 Pro-UG UG?

 MacLaughlin (1998) and Japanese to  White (1991), ESL kids coming from English via Russian anaphors. French don’t seem to learn that the verb  Kanno (1996) and JSL learners seeming to doesn’t raise (at least over ). know how to drop case markers without  Hawkins et al. (1993), FSL people seem to instruction. be “faking” French—early stage treating negation as part of the verb, start to allow SVAO in addition to SAVO (recruiting HNP shift).

L2A: Is there a difference Some things we know about between kids and adults? native languages  L2A is harder as you get older.  The differences between knowing one language and another are primarily knowing…  L1A is quite possibility bounded in time.  Different vocabulary  Evidence for CPs seem to point to different CPs for  Different roots different subsystems…  Different morphology and rules of morphological combination  Different parameter settings (perhaps in the lexicon of the language)  CPs exist in vision, maybe we can find a brain  Does the language allow null subjects? correlate?  Does the verb move to T?  Does the language allow complex onsets in its syllables?  Yet some people may manage to overcome this and  Different cultural conventions become indistinguishable from a native speaker.  Standard way to refuse, an invitation, apologize, … Some plasticity remains?  Idiomatic meanings for words and word groups  What disappears/deteriorates? UG? LAD?  Cultural literacy for metaphors and allusions  Prescriptive rules

Modeling the human L2A capacity for language  UG provides the parameters and contains the  Perhaps the LAD operates in L1A but not in grammatical system that makes use of them. adult L2A, that the language input needs to find  LAD sets the parameters based on the PLD. its way into the interlanguage some other way. Responsible for getting language to kids. intake

LAD LAD UG UG PLD NPAH Subjacency NPAH Subjacency

14 Critical period Critical period

 Lenneberg (1967). Critical periods are rampant  If exists, best candidate for cause is brain in the natural world. development.  CP for developing binocular vision in macaque  Lateralization? Maybe, but probably finished too monkeys, cats. early.  CP for imprinting in birds  Myelinization (limits plasticity)? Maybe, but probably  Delay in cataract surgery can fail to yield sight. finished too late. But maybe.  And in language-related domains too…  In the model of acquisition, what goes away?  Genie, kept from language input until 13;7  LAD?  Young kids can recover from CNS damage in ways adults seem not able to.  Plasticity in possible language knowledge (locked in place)?

Critical period L2A: Negative evidence useful?

 Johnson and Newport. Found negative correlation between age of initial exposure to language and eventual  L1A doesn’t use negative evidence. performance. Tested subjects’ judgments concerning violations of Subjacency (limits possible wh-questions,  If there is parameter transfer into IL from putative universal principle). Rapid drop-off of L1, logical subset relations might require performance after initial age around 14. negative evidence to reach correct  White and Genesee, Birdsong cite small number of late parameter setting. learners who do seem to reach a level where they are indistinguishable from native speakers.  Providing people with negative evidence seems to help—but only in the short term  So, it seems like there is at least a sensitive period, but (without prolonged practicing), it may not certain people (who work hard, care a lot, have high verbal “aptitude”?) can overcome the obstacle. yield any permanent “parameter resetting.”

L2A: Markedness? OI’s in adults? No, L2A≠L1A  Almost no finite (inflected) verb forms in non-finite contexts.  Are “unmarked” things easier/quicker to  When verbs are marked with inflection, they systematically learn than “marked” things? Does (overwhelmingly) appear before negation (i.e., they move). teaching the “marked” things give you the  Many of nonfinite forms used in finite contexts (used finitely, moved). —Prévost & White “unmarked” things for free? Oblig. Fin Oblig. Nonfin  What actually are the marked and +Fin -Fin -Fin +Fin unmarked things? (This may have more to A(F) 767 243 278 17 do with non-acquisition oriented Z(F) 755 224 156 2 theoretical linguistics) A(G) 389 45 76 7 Z(G) 434 85 98 6

15 UG access and transfer Input to intake

 To what extent do second language  For intake to work (in any kind of automatic learners know what “languages are like”? way), the data must be available. But the L1 can (Do they still know what all the potentially filter out useful information. possibilities are?)  Infants start with but lose the ability to distinguish non-native contrasts.  To what extent do second language  French “irregulars” cédez vs. cède. learners assume that the language they’re  Phonological features, distinctions, l/r in learning is like the language they already Mandarin vs. Japanese; geminates in E→J. know?

Markedness and what Markedness and what languages are like languages are like  Typological universals reduce the number of  Eckman, Moravcsik, Wirth (1989).  J/K/T→E. All wh-fronted; some had wh-inversion (wh-inv→wh- possible languages. fronting). Some yn-inv, ~all had wh-inv. Some other (wh-inv).  Marked implies unmarked (yn-inv→wh-inv).  IL seems to obey typological universals—it’s a language  having a dual implies having a plural in the relevant sense.  having purple implies having green  Markedness Differential Hypothesis (Eckman):  having wh-inversion implies having wh-fronting Difficulty in learning area of L2 from L1 if they differ  having yes-no inversion implies having wh-inversion and L2 version is more marked.  Some evidence that teaching marked structures is hard,  being able to form relatives on OPREP implies being but gives you unmarked structures for free. able to form relatives on IO

Markedness and what Language attrition languages are like  Sonority hierarchy  L1 attrition—altering L1 parameter  a > i > r > l > n > s > t settings?  Syllables as sonority waves; languages differ on steepness requirements between margin and  Null subjects: Italian speakers immersed nucleus. in English will sometime produce/accept  Most evidence that we have so far points to a big overt subjects where monolinguals would role for transfer in phonological parameters and not. Broadening the contexts in which not a lot of parameter resetting. they can use overt pronouns (not  Yet, the evidence in the phonology might be more readily available. forgetting how to use null subjects).

16 Conclusions? Bottom line:

 LAD probably atrophied (critical period): Meisel 1997.  Especially with respect to L2A, there are a lot  Universal constraints (also active in L1) constrain IL—would be true even if we were just talking about of things left to discover because careful “speaking L1 with L2 words” (Kanno 1996) and theoretically informed experiments still  L2 learners (even kids) don’t seem to set the verb movement need to be done. or null subject parameters for the target language (predicted  Many of the experiments that are in the clustering not observed) (White, Trahey, Hawkins et al.). literature rely on misleading simplistic notions (a monolithic UG subsuming the LAD,  Parameters of binding theory if correctly analyzed do seem to a single once-and-for-all CPH, a one-stage-at- be being reset. One piece of positive evidence we’ve got. a time view of acquisition, a subset relation Possibly also the Hulk results about Dutch/French. for placement or binding domain definitions)…

        

17