Proc SocAntiq Scot, (1986)6 11 , 223-253 The declining Pictish symbol - a reappraisal The late Gordon Murray SUMMARY The paper is mainly concerned with the three commonest Pictish symbols, the crescent, the double disc with Pictish the Z-rod and 'elephant' 'beast'.BStevensonR or K ideasDr The of and Dr I Henderson are outlined, namely that for each of these symbols a stylistic 'declining sequence' can be traced that corresponds approximately to a chronological sequence, enabling the probable place of origin of the symbol to be determined. The forms and distributions of the three symbols are examined in detail and it is argued that the finer examples of each are centred in different areas. For reasons which are stated, the classification of the crescent differs here from that made by Stevenson. The different decorative forms show significantly different distributions originthe but appears be to north. far most The the typicalin examples Z-rodthe of accompanying doublethe discfoundare predominantly in Aberdeenshire, where it is suggested that the symbol may have originated. Examples Pictishthe of beast hereare graded according extentthe to that their features correspond otherwiseor with lista whatof 'classical' the appear be to features form. distributionofthe The and general quality existingof examples suggest that originthe centre thisof symbol probablyis the in area Angusof easternand Perthshire. The paper also discusses arrangementthe symbolsthe of statements, in with some tentative remarks on the relative chronology of the mirror appearing alone as a qualifier. INTRODUCTION principle Th e declininth f eo g symbo thas i l t there existe prototypda r 'correcteo ' forr mfo at least some of the Pictish symbols, to which all surviving instances approximate in varying degrees, but from which later examples tend to depart more than earlier ones. This idea was StevensodevelopeK B R y nb d (1955, 104-6) with referenc decoratioe th cresceno e t e th f d no an t HendersoB applieI r D y db n Z-rods (1958d decoratioe an th Pictis - e o )V t th , e shapd hth nan f eo notchee 'beastth d an d' rectangle t seemI . s likely thadesige th teac r nfo f thes horiginalls o ewa y compled finan e underwend xan t subsequent simplification, rather than thae mosth t t complex designs result from the deliberate combination of features used separately in earlier examples. Hence in all these cases a sequence of decline in quality can be traced that corresponds approximately to a chronological sequence. This provides a method for determining roughly the relative ages of individual examples of these symbols. It follows that this sequence should help to suggest where the carving of these symbols on stone originated, the probable place of origin being where the best (by hypothesis the earliest) symbols are found. By consideration of the forms of several of these symbols, Henderson (1958, 50-2 s suggesteha ) d thae symbolisth t s develope mwa e are th a n i daroun d Invernesd an s 4 22 SOCIET ANTIQUARIEF YO SCOTLANDF SO , 1986 extendin e Golspie coasth th o t p t eu g area, constitutin e so-calledth g origin e centrth f o e symbolism . e symbol-stoneManth f o y n northeri s n Scotlan e indeear d f supero d b quality. However, when the commonest symbols are considered individually, their distributions vary markedl t wil i e argue b d l an yn thi i d s paper, fro ma detaile d examinatio e formth d f an so n distributions of the crescent, the double disc with its accompanying Z-rod, and the Pictish beast, tha finee th t r example eacf so h appea centree b o rt separatn di e areas, suggestine th l gal that no t symbol theid r snorthha rfa origie .th n ni regionaA l origi particulaa r nfo r symbol doe t necessarilsno y suppor viee th tw that certain symbols have a tribal or other regional meaning. Such a view can only be maintained if several symbol e showb n havo t nca s e regional, non-overlapping distributions e evidencTh . e doet no s seem to allow this, and in this paper attempts to seek any interpretation of the meaning of the symbols have been consciously avoided. Before considering these symbols, however, there are some cautions to be noted regarding the use of the underlying principle. The notion of the declining symbol is helpful, but there are difficulties concernin applications git . To begin with, only the commonest symbols survive in numbers sufficient for a detailed examination of their distribution or variations in design. Even in these cases, when broken down into different forms or areas, the numbers involved may be very small, so that conclusions are often tentative. Further, many of the existing symbol-bearing stones are partially damaged or defaced thao qualite s ,symbole th t th f yo s cannot alway accuratele sb y determined. The number of examples lost cannot be ascertained, but it is probable that the surviving Pictish monuments are only a small fraction of those that once existed. Hence it is unlikely that the classica r earlieso l symboy t foran f mo l survives. Eve bese nth t existing form likele e b ar s o yt no more than close approximations to the prototypes. Moreoverincorrece b y ma supposo t tt i , e tha a givet n symbol would devolv r becomo e e debased at the same time everywhere in the country. Too little is known about the means by which patterns were transmitted throughout Pictland (the symbolism being maintained wita h remarkable degree of uniformity in the process) for us to dismiss entirely the possibility of conservatis a centr n m i f excellence o e desigA . northern i n n Scotlan de supposeclosth o t e d classical form could be contemporary with a devolved design in another area, for instance Aberdeen. The majority of symbols are found on sculptured stones belonging to J Romilly Alien's Clas , undressesI d boulders with incised symbols (Alien 1903 , 3-4) II executioe , Th . f thesno e symbols appear displao st greateya r degre f restraineo regularitd an t y than those foun othen do r monument objectsd an s . This inherent conservatis r mtendenco preservo yt e stereotyped forms may indicate tha inspiratioe th t symbolise th r nfo mclosepresumes e i th o t r d exempla Clasn o r sI monuments than elsewhere. Pictish symbol monumentn so s belongin Alien'o gt s Clas I (dressesI d slabs with symbols, cros figurd san e sculptur relief)n ei , while still prominently displayee th n do reverse of the slabs, appear to be less important than those of Class I because they are subordinated to the central feature of the cross and are found alongside a variety of figures and decorative patterns. Also, the decoration and form of the symbols are generally less rigid on Class II monuments, and many, though not all, examples carved in relief show designs completely different from those tha incisee ar t stonen do eithef so r class. This chang emphasin ei s fit witn si h the likelihood that at least some Class I stones pre-date the earliest Class II stones. Other example easilo s t symbolf fi ys o t int declinine no oth o sd g sequence. Thos metan eo l objects show some affinities with relief example Clasn o s I stonesI s . Thos cavn eo e wall pebbled an s s show designs more crude and simplified than those on free-standing monuments. MURRAY DECLININE TH : G PICTISH SYMBOL 225 The arrangement of symbols in what Professor A C Thomas (1963) named statements is far more regular on Class I monuments than elsewhere. The symbols are most often found in pairs, usually one above the other. Where three symbols appear to belong together, the one that is othere e righth th f lowes usualls o si to t r mirrore o t yth wherd an , e ther foue ear r symbolse th , last two are usually the mirror and comb. The mirror by itself and mirror with the comb appear to act as qualifiers. As these arrangements are found with such regularity one may postulate a standard or classical form for statements and the presence of two different primary symbols, with r withouo qualifiera t hers i , e referre statemena s a o dt f standaro t d form Clasn I . w thersI fe e ear deviations from this form (many exception e singlar s e animal signs whic e arguablar h t no y symbols). In Class II there are a few occasions when more than two primary symbols appear on a give t certain no a monumen fac s i f no et i whethe d an t r they shoul e considereb d e on s a d non-standard statement standaro tw s a dr o ,statements standare non-standare on on s d a r dan o , d statement. Symbols found elsewhere can rarely be assigned unambiguously to statements. There are, however, identical statements of standard form on three small metal objects. On cave walls e symbolth s form larg amorphoud ean s groups e qualifier presence Th . th e symbol d th f an seo s mentioned above in qualified and unqualified statements of standard form will be discussed further. Certain modification Alien'o t s s corpus, cited her ECMS,s ea adoptede ar . Since this paper takes account of the symbols or statements as they are found-on the monuments, rather than the monuments themselves e Romath , n numerals have r conveniencefo , , been replace Arabiy db c ones and the classification of the monuments has been transferred to the symbols found on them, so that, for instance, reference is made to 'a class 1 symbol' instead of 'a symbol on a Class I stone'.