Pref. L. L. Cavalli-5S ferza Arce degli Acetari 29 Department ef Genetics Via del Pellegrine Stanferd University Medical Center Reme Stanferd, California 13th April 1971 Dear Luca, I have tried te answer the reviewer's cemments in detail. I weuld suspect that he is prebably net a prenisterian frem the leck ef detail in the cemments but I may be wreng. It is rather surprising hew much he misses the peint ef what we are trying te de. Fer ene trained in the British "histerical"” scheel ef prehistery, it is hard net te smile at the emphasis en histery by semeene whe ebvieusly dees net .khew the actual histery er in this case prehistery. Yeu sheuld leek carefully at number 4 in the enclesed cemments. 1 may net have put this in the best terms, The lines ef argument here may well be werth develeping in seme future werk, since little critical werk has been applied te this traditienal idea and it definitely ties in with the rate yuestien This is ene ef the tepics I had eriginally theught ef including in paper Ne. II. \ | I think we will need seme travel funds in the grant but this would prebably net be needed at the start, except hepefully in getting te Stanferd. My idea weuld be te try te define the specific preblems we want te deal with and the data needed ever the first year. I have friends in England whe can send xerex cepies ef the written seurces we may need. But at a certain peint it is much better te see the material and pesple directly. When I ge te England in May, I will try te see semeene (a friend ef a friend) whe has been working en the neelithic-meselithic transitien in Helland and western Germany and may have finished his thesis. I1 will alse be seeing seme friends in Helland en my way back. Thies is the area that prebably effers the best evidence in terms ef quality and ceeperatien. I have alse discevered (this prebably belengs in quetatien marks) that Italy may effer ene ef the best eppertunities fer studying the relatien between late meselithics and earliest neelithics. The U-shaped glacial river valleys such as the Adige near Trente effer unusual stratigraphic pessibilities. The scree frem the high vertical walls builds up at a fast eneugh rate that cultural layers represent shert time intervals. At a geed lecatien for eccupation, ene cen get with a little luck a whele series ef separate layers (say 3 te 5 layers per 500 years). I knew an Italian in the Museum at Trente whe is digging such s site with beth neelithic and meselithic layers. Unfertunately, the digging is a little reugh. It was enly a few weeks age when I was werking en my thesis that I suddenly reslized the vetential ef these sites situeted at the base ef the walls. It weuld be amazing if ene ef our best leng term seurces ended up being in Italy. I knew yeu have yeur eyes en Nerth africa. It seunds «s if yeu have mere than eneugh things en yeur hands fer ene spring. With best wishes. Newre abbeat In his cemments, the reviewer dees net get dewn te the centent ef the paper. Many ef his criticisms (and the straw men he tries te bring in such as "the typelegical netien ef the Neelithic") can be answered by direct reference te the paper itself. It is ef interest that ne specific questiens er ebjectiens are raised cencerning (1) the assumptiens that the medel is based en; (2) the data used; and (3) the metheds and results ef the analysis itself. All these things are side-stepped by claiming that what is called fer here is a historical view. It sheuld be fairly ebvieus that'the medel and the anslysis are net incempatible with "histerical" interpretations ef the neelithic in Eurepe. . 1. It ebscures the events ef histery by trying te smeeth them, using a pseude-scientific medel." There are a fair number ef respensible people whe would questien the use of. "yseude"” in reference te the medel. The aim ef the analysis was net te "smeeth eut" histery er prehistery, but rather te see whether er net a patterning or regul«rity existed in the available data en the early neolithic in Eurepe. The ebserved regularity -- the high cerrelatien ceefficients fer estimates ef the everall rete ef spread -~ will, ne deubt,. strike many prehisterians as "surprising". This is semething that has net been shewn befere. Hewever, the view that the authers are trying te force the neolithic inte a fermula by means ef « statistical cenjuring trick is tetally unwerranted. In a paper te fellew, «an attempt has been :nwde te estimete several regienal er lecel rates which cen be cempsred with the everall rate. The peint that the authors expect seme variatien in regienal rates is made en page 8 of the present draft. 2. Origins ef demesticatien. The mest up-te-date acceunt ef the evidence «and ideas en varieus demesticatien tepics fer the Old Werld is te be feund in the preceedings ef the 1968 Lenden cenference, which I attended. It is ebvieus that we have net fallen inte the "at-ene-time-in-ene-place fallacy". Why then is the analysis run fer feur pessible centers? Mereever, these ere net taken as centers ef demesticatien in the paper but are used as centers of diffusien frem which the spread ef relevant demesticates takes place. The feur sites selected have the virtue that relevent demesticates are decumented at them at dates earlier than their appearance in Eurepe. The reviewer mentiens "varieus plants, grains, animals". In Eurepe, we are dealing enly with three demesticated animels (Bes, pig and sheep/geat) and twe grains (wheat and barley), as mentiened in the paper. These are the relevant demesticates. Except fer the deg, there are ne serieus claims fer ether demesticated species being feund ameng the remains ef early neelithic cultures in Euvrepe. The twe grains are perhaps the mest instructive. Demesticated ferms ef wheat and barley can readily be distinguished en merphelegical greunds frem nen-demesticated ferms. The genetic "histery"' ef wheat is quite well knewn due largely te the werk ef Riley. Almest ne ene helds the pesitien that these greins were demesticated in Eurepe (either ence er repeatedly). There is ne secure palaeobetanical evidence fer the demesticatien ef these twe grains in Nerth Africa. The place where there is geed evidence is in the Near Ewst at sites such as Ali Kesh, Beidha and Jeriche, which #11 have C 14 dates that are earlier than the radiecarben dates at early neelithic sites in Eurepe. The present paper is net cencerned with trying te determine where er when er ever hew leng a peried the actual domesticatien ef verieus species teek place. What we are cencerned with is hew -- beth in terms ef rate and mede -- wheat and barley get acress Eurepe, given that they are net demesticated in Eurepe and knewing that they have been decumented in the Near East. The mevement ef these grains acress Eurepe can enly be "explained" by seme kind ef diffusien precess. 3a. "The spread was net a single process, but depended en different creps and animels."' In passing, it may be werth mentiening thet mere than ene diffusien precess is censidered in the paper. What sre these "different creps and animals"? Dees ene get certain demesticates in certain parts ef Lurepe and net in ethers? Again, we are dealing with the five demesticates listed abeve. It is these five elements that add the new dimensien te neelithic ecenemies in Eurepe. With the exception ef the deg, ene dees net get ether domesticated species. There is ne evidence fer a differential spread ef the five unless ene is talking abeut the circumpelar regien er high meuntain districts. One dees net get a barley-pig early neelithic in ene part ef Kurepe and a wheat-Bes early neslithic in anether. There is ne regien where ene ef the five gets cempletely blecked evut. Where faunal and fleral remains have been examined clesely, ene usually gets several, if net all, ef the demesticates eccurring teyether. A point that may be werth making here is that the medel (and wnalysis) dees net set @ requirement that neolithic ecenemies are all ef a uniferm cempesitien. As ene weuld expect, the actual cempesitien ef an econeny varies frem site te site but with the five demesticates representing a regularly recurring cemmen deneminater. 3b. "The dates in seuthern France are elder than three ef the feur dates frem Italy." The twe high dates frem seuthern France are much less reliable than the feur Italian dates fer the fellewing reasen. The Itelian sites are all epen settlement sites where enly neelithic layers are present. Cave sites in Italy such as Gretta della Madenna, Arene Candide and Arma di Nasine all have beth neelithic and meselithic layers. These cave sites were kept eut ef the analysis because ef suspected contaminatien (definitely the case «t Gretta della Madenna), wnich was ene ef the criteria fer selecting the sample. Direct knewledge ef unpublished meterial in Italy made this discriminatien pessible. The twe French sites ure beth cave sites. Direct knewledge here might have led te ene er beth being deleted frem the sample, altheugh this weuld be premature witheut such knewledge. Both French sites beleng én the suspect category. In additien, beth have high standard errers ef 200 years. Mere dates are needed in France, especially «at epen sites, befere ene wants te start making tee many claims here. Mest prehisterians see an Italy-te> France trend ef develepment in material culture during the early neelithic and net the reverse. 4, Envirenmental adaptetiens. This is ene ef the traditienal ways of leeking at the questien ef the spread ef the neelithic in Eurepe. Te start with, it sheuld be mentiened that, while the envirenment can influence the rate ef spread by offering greater er lesser resistance, it cannet acceunt fer hew the meelithic actually meves ferward. This appreach (environmental adaptatien) is far frem a secure ene, until we get te places like Scandinavia in Eurepe. The fellewing three peints may help explain this statement. Suppese we are censidering a spread frem nerthern Greece te the interier ef the Balkans and Central Eurepe. (1) We start by leeking at specific envirenments. What is the extent ef difference between the seils and climate at say Pelykastren en the Axies/Vardar in Greek Macedenia and at Sephia in Bulgaria er at Belgrad in Yugeslevia? On the next leg, what is the degree ef difference between cenditions at Belgrad and Munich in seuthern Germany? There are, ef ceurse, differences in all these cases but they ere far frem menumental enes as we are eften enceuraged te believe. (2) This is particularly true in the centext ef the climate at the time. Whet is the climate actually like at c. 7,000 B.P? This is the start ef the Atlantic phase ef the Pest-glacial pellen sequence in Eurepe, when average temperatures are supovesed te be 2° C warmer than they are at present (see for example, R. G. West, Pleistecene Geelegy and Bielogy, 1968). The climate ef the Balkans dreps its guard just at the time that the spread inte this urea is taking place. One ceuld almost argue here that fer a species te stay in the same climate at this time, it has te meve north. (3) What is the petential fer bielegical adaptatien er medificatien en the part ef the demesticates we are cencerned with? This is a yuestien that has usually net been auked. The climatic and envirenmental telerances ef the three species ef animals are mere than eneugh fer the jeurney te Central Eurepe. The twe grains sre mere sensitive. But en the pesitive side, they repreduce each year. A span ef 100 years gives demesticated strains ef wheat and barley 100 generatiens te underge needed miner medificatiens threugh genetic precesses and selection. Medern genetics may be able te effer the prehisterian seme help here. This is after all a genetics preblem te a censiderable extent. Censidering the rather medest adaptatiens required by envirenmental differences in relatien te the number ef generatiens invelved, the petential fer bielegical adaptatien en the part ef demesticated strains ef wheat and barley may well exceed the ebserved rate ef spread. The held up, if there is ene, is prebably net en the bielegical-envirenmental side. 5. Scandinavia. There is general agreement ameng prehisterians that envirenmental facters play a majer rele in Scandinavia. The analysis ef lecal rates, «s well as the lines in Figure 2, indicate that semething special is happening here. By c. 5,500 B.P., mest ef Eurepe belew 54° N Lat. is cevered with neelithic cultures. Scandinavia is the nerthern Eurepean terminus of the spread. Scandinavia represents a rather special case. This is clearly revealed by the Ertebglie culture which prevails in the area at ce 5,500 B.P. It is net our aim te gless ever the situatien in Scandinavia, which is breught eut by the analysis itself. At the same time, the situation here dees net undermine the validity ef the anslysis. 6-8. Starting dates. In paragraph 6, one can enly ask where the starting date ef 7,000 B.P. in Greece cemes frem. The five Greek dates listed in Figure 2 are all higher than 7,500 B.P., including the date fer Cerfu in nerth-west Greece. The "histeric" date at which the neelithic arrived in seuthern Italy is, ne deubt, prier te the Scaramella date ef 7,000 B.P.; the standard errer alene allews a date ef 7,100 B.P. lie have, as always with C 14 dates, enly appreximatiens. The mere determinatiens and the mere sites with dates the better the appreximatiens beceme. In the next paragraph, the reviewer claims that the "spread nerth started earlier", Is there geed evidence fer this? The available evidence indicates that the spread te Italy and the interier ef The Balkans is taking place at essentially the same time. It is prebably a mistake te start talking abeut starting times fer varieus spreads. The C 14 dates fer a site can be viewed much mere usefully as arrival times (er marking the date ef passage ef the wave frent), since they date eccupatien and the earliest eccupation layer at a site was used whenever dates were available fer it. One can enly use the time interval differences between sites with due cautien, fer ebvieus reasens (the standard errers). This is semething we have fertunately been able te aveid in the analysis shewn in Figure 1. The reviewer dees net seem te have perceived this advance in methedelegy.