J.A. Lakarra et Al. Euskal hiztegi historiko-etimologikoa
(EHHE), Bilbao, Euskaltzaindia, 2019.
by Michel Morvan
+ format PDF
In the year 2019 was published by J. A. Lakarra et al. a big volume with the title Euskal Hiztegi Historiko-Etimologikoa (EHHE). The aim of the book is to present a broad scope of the research about the so-called putative Proto-Basque language and the state-of-the-art of that research.
The first important research about Proto-Basque has been initiated by the prominent L. Michelena and published in 1961(1977) in his landmark book Fonética Histórica Vasca. All followers of Michelena will have to take in account his magisterial work when working on Basque. It is what J.A. Lakarra is trying to do in his numerous articles and with his good knowledge of Basque as a mother language but nevertheless with more or less success in the field of etymology as we shall see. There is a lot of ideas in the reconstruction or classification tentatives. Some are correct, but many should be clearly and definitely rejected like for instance ipurdi ‘arse’ from an impossible and untenable rather fantastic *ibi-erdi ’middle of the ford’. This is unfortunately the result of a research that bases itself exclusively on internal reconstruction, which may produce circularity and even some etymologic monsters.
The introduction.
The book begins with a very long introduction of 289 pages in Roman numerals until cclxxxix including lists and a very rich bibliography. All styles of linguistic references are used and the aim of the author seems to be the justification of his linguistic internal reconstruction. The model of what has been successful in the case of Indo-European is praised. In the mind of the author, this model can be applied to all languages and to the Basque language. But the perhaps too long introduction might be a bit boring for the reader and gives the impression of a rather big mixture of many different kinds of things and arguments without an actually clear structured development. The book presents the different researches and hypotheses about the eventual kinship of Basque that have been done. The huge bibliography is rich and very interesting and many new references of texts give the opportunity to allow more remote dates for words than the former attestations (for instance compared with the ones of the Orotariko Euskal Hiztegia). However the author does not mention in the bibliography some articles like the one of Roslyn M. Frank (journal ARSE, n°45, 2011: 17-64) which is in opposition with his own views. The mention of the book of Michel Morvan Les origines linguistiques du basque (Bordeaux, 1996) is forgotten. Only his doctoral thesis of 1992 is mentioned.
On page XV (xv) is it actually necessary to distinguish AEZ Aintzineuskara Zaharra and AEM Aintzineuskara Modernoa ? Proto-Basque is Proto-Basque. It should perhaps be also necessary to distinguish Proto-Basque and Pre-Basque. It is not exactly the same thing. Pre-Basque is not yet Basque, though we admit that Basque is so old that the distinction between Proto- and Pre-Basque is somewhat difficult.
On page XXXI (xxxvi) we agree that glottochronology has not shown a great success with the help of the Swadesh list or not.
On page XXXII (xxxii) we don’t agree with the excessive critic of J.H. Greenberg. Multicomparatism has some value. If done with enough cautiousness, it may help, especially when establishment of series is possible, which is for instance the case for words like Basque guti ‘little’ or oka ‘to vomit’ that have correspondents in Dravidian (kuti) and Uralic (Hungarian okád, Finnish okse).
On page LIV (liv) of the introduction J.A. Lakarra mentions two specialists of Uralic who have written in 2004 in the journal Finnisch Ugrische Forschungen vol. 58: 454-456 a critic recension of M. Morvan Les origines linguistiques du basque, Bordeaux, 1996, saying that there is no link between Basque and Uralic. These authors did not understand that Basque is of course not an Uralic language, but has some cognacy with some Uralic vocabulary as it has with other Eurasian families since it is a Proto-Eurasian language. They use like always the argument of the chance resemblances. They do forget or don’t know, and J.A. Lakarra also and many others, that the Basque people has a 12% or 13% component of the haplogroup U5 of the mitochondrial DNA (mtDNA). The haplogroup U5 is typical of the Uralic peoples. Thus it is not surprising to find some Uralic or Proto-Uralic words in Basque like for instance Basque suge ’snake’ and Old Estonian siug ’id.’ (from Proto-Uralic *suge ‘worm, snake’) or Basque san, sain ’nerve, sinew’ (graph zan, zain) and Mordvin san ’id.’ (Uralic *sani or *sVni ), *ne ’woman’ in Uralic and Basque ne-ska ‘girl’ with diminutive suffix -ska or the feminine suffix -n in the conjugated form -dun ‘you have, you woman’) of the second person singular (versus -duk for the masculine). If not from *anaie-ba, cf. also *ne-ba ’brother of a woman’. The very old Eurasian opposition n/k or n/t for feminine/masculine is thus very well represented in Basque. Cf. also Basque no for calling a woman versus to for calling a man, from the Eurasian alternance k /t versus n. As a Proto-Eurasian language, Basque is cognate with more than one particular Eurasian family. For ’woman’ Uralic has *ne, Finnish has nieid, nais, neis, nis, naine, Sami has nis, niso, nieida, Hungarian has nő, Sumerian has nu, Chinese has nü- Koriak has ñe. Cf. also for instance Proto-Basque *oka ’to belch, vomit’ and Uralic *ok- ’id.’, Finnish/Estonian okse ’id.’, Hungarian okád ’id.’ In Altaic: Turkish ög- , ök- ’id.’, Mongolian og-, ogi ’id.’, Tungus-Manchu okor- ’id.’. It cannot be of course a chance resemblance. The fact that it may be an expressive word is not a hindrance. Many prehistoric words are of expressive origin. The same is valid for Basque khe ’smoke’ for instance, an old Proto-Eurasian word cognate with some Caucasian forms with the meaning ’cough, smoke’ and even until some Amerindian languages (Hokan qhe ’smoke’ in C. Jany, 2009, p. 49). It cannot come from *gene > *gee as given by the author.
On page LXXXI (lxxxi) we can see how J.A. Lakarra lets come many completely different words like edin, jaio, jarrain from a unique root *din. This way to do is again a consequence of the fact to rely exclusively on the internal reconstruction. Internal reconstruction is of course necessary but not with so much excess. There are other similar biases for instance with *bar or gain/gan as we shall see.
On page LXXXV (lxxxv) appears the unfortunately untenable etymology ipurdi ’arse’ from *ibi-erdi ’middle of the ford’ ! The right analysis is of course *i-purd-i like for many other Basque words the stem of which is embedded within a prefix i- (or e-) and a suffix – i. Cp. *i-dur-i > iduri ‘appearance’. The Pre-Basque/ Eurasian expressive stem *pVrd means also ’fart’. We should note that the Basque word eper ’partridge’ is known also as eperdikara ’trembling of the arse’. The Basque word is not a borrowing from Latin perdix, but the Latin word comes also probably from the very old Eurasian stem *pVrd.
The Dictionary.
The Dictionary does not follow exactly the alphabetical order since J.A. Lakarra gives after each entry the words that he thinks to be derived from the same stem. This makes the reading rather difficult.
Page 4. abar ’twig, branch’. Is analyzed as *a-bar with *bar meaning behe, hondo ’down, bottom’ (?) as in ibar ’valley’ or barren ’’inside’’ (?). The word abar has to be linked to abe ’wood, post, support’ and has possibly undergone the influence of adar ’branch’. The author bases his etymology only on the segment bar of abar which is a rather poor argument for a putative comparison with ibar or barren. There is no semantic link between ’twig, branch’ and ’down, bottom’. The only external shape of a phonetic segment is not sufficient for a correct comparison. This is one more time one of the flaws of the want to rely only on internal reconstruction under the pretext that Basque would be an isolate. But isolate does not mean that there is no links with other Pre-Indo-European languages.
Page 5. abadia. Why is the word followed by the date of 1562 Land. (Landuchio) ? It is very late. The borrowed Roman word is much older. Is there actually a good reason to distinguish the dates of the Roman or Middle Ages time from more recent ones even if the book is called Historiko-Etimologikoa ?
Page 5. abaritz ’holm oak’ (Spanish carrasco). The following date 1765 Lar. is too late since the word is attested in 1205 (Guillermus de Avariç). Is it necessary to put 1765 Lar. (Larramendi) after the entry?
Page 5. abaro (= zuhaitzpe) ’refuge for the cattle’ 1399 .The word is probably older since it is a derivate from abar ’branch’. It is not very important.
Page 8. abendu ’December’. 1496 RS cannot be understood. The Refranes (RS) are dated from 1596, not from 1496. The author has taken the idea of putitng the dates one century in advance from the thesis of Manuel Padilla about the souletin dialect (2017) where Oihenart is dated from 1567 instead of 1657. Why not two centuries, three centuries, four centuriess in advance and so on?
Page 9. abere ’animal’. 1496 RS for Refranes 1596. The word is much older. Cf. 1193 Auereco. Is it necessary to put the date of the Refranes after the word ?
Page 14/15. jarri ’sat’.The analyze *e-arr-i is correct but not the following *arr < *Car or *arr < *dar or even *bar that are arbitrary.
Page 15. lar ’too much’ ( = gehiegi). 1496 RS for Refranes 1596. From lar <*dar and adar ? The link with lahar ’bramble’ and the link with larre ’heath’ are too dubious. The sole outside or external shape of words is of course not sufficient at all. Generally this way to do is rather the one of amateurs. There is also no real evidence that the original meaning should be hazi, hazkor ’grown, growing strong’.
Page 16. larri ’severe, excessive’. 1527 Zalgiz (de Sauguis) and larritu 1496 RS. The traditional date for Zalgiz (Bertrand de Sauguis) is 1627. The date for Refranes is 1596.
Page 16/17. larre ’heath’ 869. The cognacy with adar ’branch, horn’ from a root *dar is absolutely untenable.
Page 17. larratz ’cog’. Of course no link at all with adar ’branch, horn’ (+ wordlist p. 715).
Page 18. larrain ’area’. Of course no link at all with adar ’branch, horn’ (+ wordlist p. 715) . The word is a compound *larre-gain.
page 22. agur ’salute, hello’. An etymology from Basque gur ’bent, reverence’ is better than the unsure one from Latin augurium ’omen’.
Page 23. ahaide (aide) ’parent’.1496 RS for Refranes 1596. The etymology from anai ’brother’ is unsure because it would be strange that a so basic word could have lost its intervocalic consonant n , especially when there are some attested strong forms annai with geminate. Perhaps from aho-ide.
Page 24/25. aho (1173, ao) ’mouth’. From *ano < *a-no. Very doubtful. Certainly not from ate ’door’ and -n
Page 31-33 aita ’father’ 991. The old prehistoric word is attested in whole Eurasia (Turkish ata) until the Paleo-Siberian Ghiliak (Nivkh) ïtïx. In Aquitanian atta. Of course J.A. Lakarra does not mention the inscription ATA of Veleia. He does not accept as authentical the inscriptions of Veleia (IVth-VIth centuries). But in these inscriptions we have not aita but ata, the older form of aita : IOSHE ATA TA MIRIAN AMA. It might be a rather good argument for the authenticity of the inscriptions.
Page 35. aitatu ’mentioned’ 1496 RS for Refranes 1596 and aipatu (aipha-) 1527 Zalgiz for de Sauguis 1627. The original form is aipatu from aipa.
Page 36. aizkora ’axe’. From Latin asciola ’little axe’ or from (h)aitz ’stone’ like in other names of tools. The Basque forms with h- are not always etymological, far from it, what many authors have forgotten. The word hezkabia ’itch disease’ (Latin scabies) is not crossed with hatz ’finger’ as J.A. Lakarra writes. It is a borrowing from Roman escabia (Latin scabies). The h- is only not etymological like in harma ’weapon’ for Roman arma. Many scholars did not see that there was an imitative or regressive systematization of h- in tne northern dialects. It is a pitfall for those who don’t know the reality and the subtleties of the Basque language.
Page 37. alhatu ’given food, pastured’ 1527 Zalgiz for de Sauguis 1627. The etymology from Germanic *halon wherefrom Roman *hala (French haler ) ’to pull, to haul, to tow’ does not fit for the meaning ’to give food, to eat, to have pasture’ (the meaning ’to itch’ being secondary, cp. French manger and démanger, wherefrom the derivate meaning ’to torment’). The form ala ’to feed, pasture’ seems to be attested in some Celtic or Germanic languages and perhaps even in Low Latin or Vulgar Latin.
Page 40. halatu ’towed’ 1725. Probable confusion with hala, ala ’flat boat’ and of course complete confusion with alha ’pasture’. The word ala ’flat boat’, the meaning of which is also ’oar’, is more probably derived from Roman or Spanish ala ’wing’. The initial h- is not always etymological in the northern dialects.
Page 41. alhaba ’daughter’. The right etymology is of course alu ’vagina’ (ala- in compounds) + kinship suffix -ba.
Page 42/43. ama ’mother’ 1028 San Millán. There is no known reason for a link with the toponymic suffix -ama in Zegama, Ultzama, etc. The word ama is of course much older than 1028. It appears in the inscriptions of Veleia : IOSHE ATA TA MIRIAN AMA. It is a very old prehistoric word known in whole Eurasia : Sumerian ama ’mother’, Yeniseian ama ’mother’, Paleo-Asiatic Ghiliak (Nivkh) ïmï-x, etc. It is not a recent Wanderwort even if originally expressive. A very old cognacy with Eurasian *eme ’woman’ is possible. Cp. Finnish emä ’mother’.
Page 44. amona ’grandmother’ 1745 Larramendi. Not necessary. There is Amunna in San Millán (year 759).
Page 45. amets ’dream’ 1527 Zalgiz for de Sauguis 1627. From ambets ? Absolutely untenable and fantastic hypothesis *gogoan-ets > *owanets > *owahets > *owaets > *awaets > *awaents > *abents > *ame(n)ts > *amets (sic). This is not actually etymological science. Only a long chain with each time slightly changes giving opportunity to obtain anything one wants to obtain.
Page 46. anaia ’brother’. The variant annai with geminate N is an argument against the analyze of ahaide from anai. It would be strange that a so basic word would have lost its intervocalic n. The word is Proto-Eurasian: Dravidian anna ’brother’, Turkish anai ’brother, uncle’, Japanese ani ’brother’, Inuit ani ’brother’, Navajo anaai ’brother’. Possibly built on the same stem as Eurasian *ana ’mother’. Cp. for instance Turkish ana ’mother’, Tagalog nanay ’mother’ with reduplication.
Page 48. neba. ’brother of a woman’. Not from ama ’mother’ or anai(e) ’brother’. From Proto-Basque *ne ’woman’ and kinship suffix -ba. Uralic has *ne ’woman’. Basque has also neska ’girl’ (probably from *ne-ska with diminutive suffix -ska) and the mark n for the feminine (k or t for the masculine) at the second person of the singular: -dun < -*du-n. Cf. also no for calling a woman (in opposition with to for a man).
Page 51. apal ‘low’. The etymology *adval < Latin ad vallem is dubious. Cf. zapal. Some scholars have said the same for the dubious etymology ibar ’valley’ < Latin in vallem.
Page 52. zapal ’flat’. Not from the wrong analyze *za-pal. What is the mysterious za-?. The right stem is *zap- ’flat’ (*sap-) with adjectival suffix -al.
Page 55. ate ’door’. Would come from *a-te. Very unsure. The cutting may be arbitrary.
Page 59. zaldar ’boil, furuncle’ 1496 RS. for Refranes 1596. The end -dar might perhaps be adar ’horn’ or only a nominal suffix -ar.
Page 61. zaldi ’horse’. For J.A. Lakarra from *zal- di ? Perhaps possible because of zaldun ’horse rider, knight’ but only if the latter is not a contracted form *zaldi-dun.
Page 64. bake ’peace’ 1496 RS. for Refranes 1596. The word is of course much older since it is a loanword from Latin pace(m).
Page 68/69. baratze ’garden’. R. L. Trask links it with baratu ’gelditu’, but it is not very convincing. In the present case, baratu could only mean ’closed’. It also does not come from barat (Corominas 1972) ’hole, ditch’ (zulo, erreten) and also not from vallis ’valley’ or Roman prat ’meadow’. The good etymology is probably *bara-tze with bara ’bar, fence’ (enclosure) and nominal or deverbal suffix -tze.
Page 70/71. bare 2. ’slug’ 1496 RS. for Refranes 1596. The word for slug is much older. It cannot be separated from bare 1 ’slow’ and bare 3 ’spleen’. Already attested in 947.
Page 71. baraistu (Azkue) ’patience’. Perhaps from *bar- or bare but certainly not cognate with barren, barru ’inside’. Why should we rely only on the outside or external shape of words with so different meanings ? Only because of a segment bar? All words that contain bar cannot go back to the same root *bar-. It is a very naive concept. The stem bare has a simple r, the stem of barru (*baRu) a strong rr.
Page 74/75. bare 3. ’spleen’ 1496 RS. for Refranes 1596. The word is linked with bare 1 ’slow’ and with bare 2 ’slug’. It has nothing to do with *barr-, barren, barru ’inside’ as mentioned. Only because of a similar segment bar ? Absolutely untenable.
Page 75. bazkari ’meal’. There is no obvious link between bazkari ’meal’ and baratu ’prepared’ (Roman parar) in spite of the variant barazkari. The word bazkari ’meal’ is very probably related to bazka ’to eat, pasture’ and the form barazkari possibly crossed with baratze ’garden’.
Page 77/78. barren ’inside’ 1049. De *bar-hen. Why -hen ? The stem is *barr- and the suffix is -en. And what is the link with bare, are ’slug’ ? There is none.The link between barren and ibar ’valley’ remains also unsure.
Page 80. barrun ’inside’. For Michelena from *bar + un (une) ’place’. And for J.A. Lakarra from bar + *bun, buno (sic).
Page 81. barre ’to laugh’ 1496 RS. for Refranes 1596. Again from *bar ’down, bottom’ (behe, hondo) ! Only because of a common segment bar. Of course untenable.
Page 82. bartz ’louse’ 1557 Oih. Atsot. for Oihenart 1657. Again analyzed as *bar-tz from the alleged unique root *bar ! It is unsound.
Page 83. baso ’wood, forest’. Again from *bar-so < *bar- ’behe, hondo’. Not convincing. The examples of ürsan (S) > usai ’sneezing’ or mersede > mesede are not sufficient to prove anything in this case. The want to derive so many words from a unique common stem *bar- is untenable. It is a bad side effect of the ’all explained’ with the help of internal reconstruction. It is complete circularity.
Page 85. bat ’one’. From *bade, bede (cf. bederatzi ’nine’). From a curious *ba-de ? Dubious.
Page 84. bazter ’side apart’ 1070. From *barzter and even *barz-der. Dubious. And again from *bar- ’behe, hondo’. The want to let come so many completely different words (barre ’laugh’, bartz ’louse’, baso ’wood, forest’, bazter, etc.) to only one and same root *bar- is extremely strange and of course in no way convincing. It is obviously a very excessive use of internal reconstruction based only on the partial external shape of words. It has no future.
Page 92. beti ’always’ 1490 GaribAtsot. for 1590 Refranes de Garibay. From *be-ti (?).
Page 94. bage ’without’. For J.A. Lakarra from bat ’one’ + ge (?). Only -ge would be the privative. This is only possible if -ge is not itself a contracted form of bage. And there is the much used variant gabe.
Page 95. begi ’eye’ 1080. Attested in 984. So why 1080 ?
Page 97. begi ’eye’. Still unsure for *b-egi with *b- as body class prefix or *b- for bi ’two’. Cp. egun ’day’ (*eg-un), eki, eguzki ’sun’, egia ’truth’ (?). Would be the same as in belaun ’knee’, beso ’arm’, belarri ’ear’, etc. In some languages a prefix b- or similar seems to exist with body parts or other things as a possessive of third person, so that for instance a word like begi would be to analyze as ’his eye’. This possessive hypothesis had been mentioned by C.C. Uhlenbeck in 1927. But many other Basque body part names have no prefix *b-. And what about buru ’head’ ? Or belar ’forehead’ ? For these latter ones a prefix bi ’two’ is not possible. For some of them like belaun ’knee’, bular ’chest’ or belar ’forehead’ we would suspect an original Proto-Basque stem *bVl- meaning ’rounded shape’.
Page 98. begiratu ’observed’ 1527 Zalgiz for de Sauguis 1627. But this verb is already present by Echepare 1545. Would come from Latin uigilare. Not impossible but difficult because of begi ’eye’. It might perhaps be from begi ’eye’ and adlatve suffix -ra.
Page 100. beharri ’ear’ 1557 Oih.Atsot. for Oihenart 1657. Probably from beha. The form belarri might be a variant due to a crossing with belar ’forehead’ or influenced by it. Cf. the verb behatu.
Page 104. belatz ’crow’ 1235. Why put 1235 since there is already Belaza in 950 ? The date 1235 is also from the Middle Ages. To separate the date of the entry of those of the Middle Ages is perhaps somewhat artificial.
Page 107. berna ’leg’ 1496 RS. for Refranes 1596. Idem for bernazaki. What is -zaki ? The meaning is probably ’bone’. But for J.A. Lakarra from *zarki ’haragi zahar’ (old meat) ? Of course untenable. What is the semantic link of ’old meat’ with the meaning ’bone’ ? Old matter ? Unconvincing. The word zaki might perhaps be cognate with zakil /makil ’stick’. Cf. zulazaki (1745) ’palito para agujerar el pan’.
Page 108. hezur ’bone’ (azur). The Proto-Basque reconstruction *enazur for ’bone’ (Basque ezur) has been proposed by L. Michelena. It is probably wrong. The presence of the nasalized initial vowel e- in the reconstruction *enazur is due to the nazalized variant enzur of the Roncalese dialect. We give hereby the arguments against this reconstruction:
a) The nasalization of the Roncalese form might be secondary.
b) The presence of the vowel a between n and z is not justified. The interconsonantical insertion of the vowel a in the reconstruction *enazur is probably a bias due to the influence of the Biscaian variant form azur for ’bone’. It should not be forgotten that Biscaian has a variant a for e like in berri > barri (B) ’new’, berdin > bardin (B) ’similar’. It is also valid at the initial: ezur > azur.
From the Proto-Basque reconstruction *enazur which is very probably false, the author J.A. Lakarra has built an even more remote and also false Proto-Basque reconstruction *bernazur ’wood of the leg’ or ’bone of the leg’ based on the argument that in some other languages the word for bone and the word for leg are issued from the same stem like for instance in Germanic languages, wherefrom English bone ’bone’ and German Bein ’leg’. The argument might have been good but it does not fit for Basque since *bernazur cannot be older than the extremely dubious *enazur and of course ezur itself. The Basque word berna or perna ’leg’ is a Roman borrowing (wherefrom Spanish pierna) and the Basque word ezur is a very old prehistoric word. Moreover there is absolutely no justification for an evolution bern- > ben- in this case. The Basque word ezur is valid for all bones. The word bernezur or bernazur ’tibia’ exists now in modern Basque but only since the XVIIIth century (Larramendi 1745). On the contrary the word ezur ‘bone’ is attested since 1200 (Garcia Ezurra). The want to apply Indo-European semantic models to Basque may be sometimes misleading.
It should be also noticed that in Basque the initial h- (as in the variant hezur) is actually old and only remaining in the northern dialects but is also in many cases a big pitfall for the linguists because the h- is not always etymological, far from it. The initial h- in northern Basque, even if old, underwent a systemization. Many words who had no reason to begin with an initial h- have got one by a regressive analogy under the influence of other words with a possible etymological initial h-. For instance a word like the Roman borrowing arma ’weapon’ has a variant harma or arena ’sand’ which gives harea, aize ’wind’ has a variant haize although it is very probably itself a variant of the Roman borrowing aire ’air’ (Spanish aire) . The same is probably valid for genuine Basque words and this makes the case very difficult. To know if a Basque word has an etymological initial h- or not is therefore a huge and very difficult puzzle and brain-teaser. This problem is also valid for many interior h , intervocalic or not, except when the case is obviously clear: ahate < *anhate < Roman anate ’duck’, bahe ’sieve’ < *banhe < Roman vane, *mihi < Proto-Basque *bini ’tongue’, the latter being still present in the word galbin ‘clapper of a bell’.
Page 109. belaun ’knee’ 1496 RS. for Refranes 1596. From *bela-bun (sic). Or even *berna-bun (sic). Only *bela-un(e) would be acceptable. We can see here that the so-called body-part prefix b- probably does not work since we would have a strange *elaun. In fact the stem is probably *bel- with the meaning ’rounded shape’ like in belar ’forehead’ from a Proto-Basque or Pre-Basque stem *bVl-. Cf. also bular ’breast’. For the meaning cp. Tsezian gurtu ’knee’ from *gur ’round’, Bedauye kumba ’’knee’’ from *kumb ’round’, Sheka k’um, Sino-Tibetan *gup ‘curved’. In Altaic perhaps Buriat belhen ’pastern’, Kalmuk belwn-ceg ’pastern’, Udighe bäluga ’knee’, in Uralic *pol-w- (?) ’knee’ only if *pol- means ’rounded form’.
Page 112. bizi ’life’. Is it the same thing as bitxi 1330 ? From *bis-i. And is bitxi in aitabitxi, amabitxi the same word ? It is very strange. Why should it be cognate with bizi ’life’ ? Is Altaic *bisi ’life’ a lookalike (especially Tungusic) ?
Page 113. bizkor ’lively’ 1626. From *biz-kor, from *biz-i ’life’. No link with Gascon biscòr ’oblique, slanted’.
Page 116/117. bortz/ bost ’five’. The link with a putative *borts ’hand, five fingers’ (cp. Austronesian lima, rima ’five, hand’ ) is possible, but hardly with *bor ’round’: borobil, enbor ’trunk’ and let alone an untenable ortzi ’sky’ from a fantastic *bortz-i (sic).
Page 118. zazpi ’seven’. Some scholars have reconstructed a form *bortzaz-bi ’five + two’. It remains very unsure. The segment -az has no solid justification (hatz ’fingers’ ?). The probability of a cognacy with Latin septe(m) or a similar form remains very high. Cf. for instance Romanian sefti and Coptic shashfi. The same is valid for sei ’six’, wrongly derived by the author from *sehi < *seni. The word sei is a Roamn loanword even if there is no final -s in the Basque form. Cp. Italian sei ’six’.
Page 119. hogei ’twenty’ 1415. From *(b)or-gen-i > *oh-gehi > *ogehi > hogei (sic). Probably too complicate. The comparison with Celtic ugent remains possible.
Page 120. ortzi ’sky’ 1110. The word would come from *bortz- (sic) which is an overrly doubtful hypothesis, especially when applied to ortzegun ’Thursday’, the ‘day of the sky’.
Page 122. ortzi ’sky, God’ ? From *bortz (sic) < *bor ’round’ like in borobil or enbor. Absolutely untenable. The other hypothesis Urcia from hur ’water’ (rather ur) is a little bit better.
Page 123. ortzegun ’Thursday’. The hypothesis that it would come from *bortzegun ’fifth day of the week’ is completely wrong. The word is clearly ‘day of the sky’ corresponding to the Jovis dies of the Romains and Jupiter of the Greeks (cf. Spanish jueves). Moreover ortzegun is the fourth day of the week, not the fifth day (astelehen is ’monday’, first day of the week). Absolutely untenable, all the more so as J.A. Lakarra lets come ortzi ’sky’ from bortz ’five’ (sic).
Page 130. dolu ’pain, mourn, claim’ 1627 de Sauguis. But the word is already attested in 1203.
Page 133. eban /ebaki ’cut’ 1393. ebagi 1496 RS. for Refranes 1596. The stem is *bak or *ba. If the stem is *bak , there is no suffix -ki, only the participial suffix -i. It is a bit difficult, though not impossible, to let it come from a stem *ban. It is the problem of the exact status of the ending -n in some verbs like edan ’to drink, drunk’, jan ’to eat, eaten’. If the final -n does not belong to the stem, in this case the stems would be *da for edan or *a for jan.
Page 134/135. epai ’cut’. Would come from a curious form *ebanite. Oihenart 1657 has epatu from epe ’space of time’.The form epai speaks rather for a contraction of *epaki /ebaki.
Page 137. edan. ’to drink, drunk’ 1496 RS. for Refranes 1596. The verb would be cognate with ardo < *ardano ’wine’ (?). Dubious. Did the first Basques drink only wine ?
Page 139. ardo ’wine’. The reconstruction *ardano is correct since we have ardantza ’vine’ and variants like ardao. For J. A. Lakarra it would be the same stem *dan as for the verb edan ’to drink’. But if so, how to explain ar- ? And did the Proto-Basques drink only wine ? Dubious.
Page 140/141. eder ’beautiful’ 869. From *deder < *der ? The stem is actually *der. J. A. Lakarra writes that the meaning of the root *der is not obvious. We can see here once again that the internal reconstruction is not sufficient. It is necessary to make comparisons with other non-Indo-European languages of Eurasia: Hungarian der-ül ’to shine’, Ostiak eder, ater ’fine weather’, Dravidian ter ’to shine’, Japanese ter-asu ’to shine’.
Page 143. ediren/idoro ’found’ 1490 Garib.Atsot. for 1590 Refranes de Garibay. edaraite 1496 RS. for Refranes 1596. From *e-ra-din ?
Page 156/157. jaun ’sir, lord’. Would come from *e-da-dun ? And linked with edun, dun ’to have’ ? Unsure. jaube 1496 RS. for Refranes 1596.
Page 160. egosi ’boiled, cooked’ 1557 Oih. Atsot. for Oihenart 1657. For J.A. Lakarra from *e-gos-i (correct) < *egorzte < *e-gor-z-i (?). Unsure. The stem is *gos.
Page 162. jauzi ’jumped’ 1571 Leiz. Would come from *e-da-goz-i ? Unsure.
Page 163/164. urgatzi (urgazi) ’helped’ 1496 RS for Refranes 1596. For J.A. Lakarra from *e-da-ra-goz-i. Seems too complicated.
Page 168. engoitik ’already, (from) now on’ 1545. Would come from *egun-goitik ? Unsure. The segment en- might be only a Roman prefix.
Page 169. higuin ’digust’ 1490 Garib. Atsot. for Garibay 1590. higu 1557 Oih.Atsot. for Oihenart 1657. Would come from *eguni > *eguhi > *hegui. What is the semantic link of higuin ’disgust’ with egun ’day’ ?
Page 170. uda ’summer’ 1496 RS for Refranes 1596. From euda ? Dubious.
Page 174/175. gurdi ’cart’ For J.A. Lakarra from *egur-di. The problem is that the word egur is the word for the heating wood, not for the construction wood.
Page 177. ehe ’washing water’ 1557 Oih. Atsot. for Oihenart 1657. Might be from erre ’boiled’. In the Souletin dialect can happen sometimes a double loss of rr : erre > ere > ehe. Cp. for instance arrats ’evening’ > arats > ats.
Page 177. ile ’hair’ ule 1496 RS for Refranes 1596. Because of the variant ule many scholars have believed that the word was cognate with Indo-European *wul ’wool’. Dubious.
Paga 179. ile ’hair’. It is probably a mistake to believe that the variant ule would be directly cognate with Indo-European *wul- ’wool’. It should be proven that ule is the primitive form and the same for other words like iri / uri ’domain, city’. Cp. perhaps Dravidian ile ’hair’ ?
Page 183. ero 1 ’kill, death’ 1406. Would come from *e-ra-non ? Very unsure.
Page 185. haragi ’meat’. From *erhagi > *arhagi ? From ero ’to kill’ ? Unsure.
Page 185. arakondo (B, Azkue) ’nudo de árbol’ (tree knot). What is the semantic link with aragi ’meat’ ? Cf. rather arakil ’branch’.
Page 189. eraman /eroan ’taken, let go’, ezteramana 1527 Zalgiz for de Sauguis 1627. The verb eroan is the factitive of joan (stem *oa) and eraman the one of eman (stem *ma). Are the two stems cognate ? Too dubious.
Page 190. ekin ’to do’ (akio) 1490 Garib.Atsot. for Garibay 1590. Cannot be separated from egin. No obvious link with jakin.
Page 192. jakin ’to know’. Why should it have a link with ekin ? Untenable.The root of jakin is *aki and the root of ekin /egin ’to do’ is *ki / *gi.
Page 198. elur ‘snow’ from *euri-hur ’rain water’ > *erhur ?. Not convincing. The word elur cannot be separated from lur ’earth’. Cp. zur ’wood’ and ezur ’bone’. They are prehistoric doublets.
Page 200. maite ’to love, beloved, dear’ 1527 Zalgiz. for de Sauguis 1627. For J.A. Lakarra from *emaite < eman ’to give’ ? Unsure.
Page 203. eme ’female’. The etymology from Latin femina > Occitan /Gascon. hemne is dubious for a so fundamental basic word. Like for its antonym ar ’male’ (Turkish är, Chuvash ar, Ghiliak ar) the word eme ’woman, female’ is present in the whole Eurasia.
Page 206. ezagun (ezaun) ’to know’ 1490 Garib. Atsot. for Garibay 1590. The link with entzun ’to hear’ is dubious.
Page 209. zorne ’pus’ 1635, zaurne 1650. The link with zauri ’wound’ is right. The word zaurne ’pus’ is a compound of zauri ’wound’ and esne ’milk’.
Page 210. erraz ’easy’ 1496 RS. for Refranes 1596. What is the link with erre ’to burn’ (p. 709) ?? Unsound.
Page 221-224. gain/gan ’on, top’. Certainly not the stem of gari ’wheat’, garbi ’tidy’, garo ’fern’, garo ’dew’ or galdu ’lost’, galde ’question’. Why should all these words go back to a unique stem *ga(r)- /*gan ? It is due to the circularity of the internal reconstruction. It is obviously nonsensical and a big methodological bias. The word gari ’wheat’ comes from a stem *gar- ’cereal, grain’ very well represented in the Near Eastern languages. It goes back to the first farmers of the Neolithic. Cp. Basque garagar ’barley’ from *gar with reduplication, Hebrew gargar ’grain’, Armenian gari ’barley’, Kurdish garis ’millet’, etc. An old stem *gar possibly reduced to zero degree *gr- in some Indo-European languages.
Page 234. gantz ’grease’. A link with gatz ’salt’ is possible. But to rely gantz with gain ’on, up, top’ is highly dubious. As for gatz, it might perhaps be compared with some Eurasian words with the form *kas- or *gas-: cp. perhaps Mongolian gasi ’bitter’ ?
Page 242. garo ’fern’. Like gari, garbi, galdu (sic) the word garo would come back to a unique stem gain/gan ’on, up’. Absolutely untenable. The same is valid for garo ’dew’.
Page 243/ 244 gari ’wheat’. From *gar-i that would be a participe with the meaning ’what has been removed from the top’ (gainazal kendu) and therefore from gain ’top’. Such an etymology is impossible and very tortuous. This is again the result of the want to explain all genuine Basque words only with the help of the internal reconstruction. The stem *gar- ’grain, cereal’ is broadly attested in the Near East. Cf. Hebrew gargar ’grain’, Armenian gari ’barley’, Kurdish garis ’millet’, etc. It goes back at least to the first farmers of the Neolithic time. With reduplication also Basque garagar ’barley’.
Page 248. garbi ’tidy, neat’ 945. For J.A. Lakarra once again from *gar- < *gain = gainazal kendu, like for gari ’wheat’ !! Absolutely untenable.
Page 251. galdu ’lost’. The verb is put in the same group as gari ’wheat’, garbi ’tidy’, garo ’fern’ from *gar < gain ’on, up’. Absolutely untenable.
Page 267. bigel ’liver’ 1490 Garib. Atsot. for Garibay 1590. The word gibel is analyzed *gi-bel. It could be correct for the meaning ’liver’ (black matter) but gibel ’back’ might be also *gib-el for instance if it should not be the same word, perhaps from *gib- ’rounded form, hump’.
Page 270. izurde ’delphin’. From giz- ’man’ and urde ’pig’. But it may be also iz ’water, sea’ and urde ’pig’. Cf. izpazter ’seaside’ < *iz-bazter.
Page 273. gorri ’red’ 1027. From *gor ’hard’ ? Unsure. Cf. also gori.
Page 282. gudu ’war’. Perhaps from German *gund(ea) (Uhlenbeck) ? Or from cudut > cudu ? Not very convincing. Iberian inscription gudua deisdea. The variant guda is secondary. But cf. for instance Dravidian guddu ’fight, quarrel’ ( Burrow /Emeneau, Dravidian Etymological Dictionary, n° 1850).
Page 283/284. gutun/kutun ’letter’. The right etymology is Roman coton, not the Arabic plural kutub from kitab ’letter’. The cotton was used as pad and support on which it was possible to write and it is why the word means also ’amulet’ because there was a tradition to put a propitiatory word in the little bell of the leading ewe of the herd.
Page 296. handi ’big’. From *han-di. Possible. Not from Latin grandis.
Page 298. anitz ’much, many’ 1557 Oih. Atsot. for Oihenart 1657. Would come from *han- as in handi ’big’. Perhaps.
Page 298/299. hamar ’ten’ 1110. From *han-bor or *han-bar ? But cf. amai ’limit’ obviously derived from the same stem.
Page 304. aker ’he-goat’. For J.A. Lakarra from *a-kher. What is *kher ? What is a-? Tne word cannot be separated from aketz ’boar’. So that the stem is probably *ake- with the meaning ’male animal’. Cf. beh-i ’cow’ and beh-or ’mare’ for the females, two words based on the same Proto-Basque stem *beh- ’female animal’ with two different suffixes. The double case of aker and aketz finds a parallel system in some Indo-European languages with Latin caper ’he-goat’ and Greek kapros ’wild boar’ from a common stem *kap(e)r-.
Page 309. aketz ’boar’ 1173 de *han-ketz (-getz) ? But cf. aker. Both words aketz ’boar’ and aker ’he-goat’ are derived from a stem *ake- ’male animal’. Cp. behi ’cow’ and behor ’mare’ (from Proto-Basque *beh- ) or ardi ’ewe’ and ahardi ’sow’ for females.
Page 317. hausnar ’rumination’ 1643. It may be only from esne ’milk’ since there is a form esnar.
Page 322. atso ’old woman’ 1496 RS. for Refranes 1596. The word cannot be separated from aso ’old’ and asaba ’ancestor’.
Page 346. hezkur ’acorn’ 1007 Ezcurra. The word comes probably from *esk- with the meaning ’tree, oak’. Why should it come from hezi ? Moreover there is no evidence that the initial h- of hezkur is etymological. On the contrary the word is always written ezkur. The form hezkur appears only one time in 1908 !
Page 347. ezker ’left’ 979. From *hez-ker ? Unsure.
Page 351. jaulki ’gone out’. Would come from Proto-Basque *e-da-dul-ki ? Unsure.
Page 351. heztul ’cough’. How to believe that the word could be cognate with hezi ’tamed, educated’ ? From heze ’humid’ ? Untenable.
Page 352. ezti ’honey’ 1330. From heze ’humid’ ? Perhaps.
Page 365/366. alanor, ohalano (Pouvreau, XVIIth c.), alandor (S). From alan and hor ’dog’. The word alan is probably a loanword from Spanish alano. It means ’dog of the Alans’ (French Alains), a mountain people of the Antiquity around the Caucasus and Northern Iran. Cf. perhaps also Abkhaz ala ’dog’. Not necessary from German *alan ’hazi’.
Page 367. hortz ’tooth’ 1217. Supposedly from Proto-Basque *hor-tz where hor would be ’dog’. Dubious.
Page 368. otso ’wolf’ from *hortz-so ’big tooth’ itself from hor ’dog’. Very unsure. The dog is issued from the wolf, not at reverse. A form with -r- never appears with that word.
Page 370. ogi ’wheat, bread’. From *hor-gi ’matter for dog’ ? The hypothesis is based on the hapax form that appears in Aymeri Picaud in the XIIth century: orgui. All other writers have ogi. Now the question is to know if the form of Picaud is right or only a mistake.
Page 379. idor ’dry’. Would be cognate with the c omplete opposite hur ’water’ (sic). Untenable.
Page 386. hibai/ ibai ’river’. From *hur-ban-i ’cut water’ ? For J. A. Lakarra the word (h)ur ’water’ would have given i- in some words like ibai ’river’ or ibar ’valley’. A bit strange. In the old toponymy the word ibai appears only as bai like in Baigorri or Baiona. And what means ’cut water’ ?
Page 392. ipurdi ’arse’ 1496 RS. for Refranes 1596. The Basque word ipurdi ’arse’ is built on the regular form of many Basque substantives embedded between two vowels i- or e- at the beginning of the word and -i at the end. These vowels don’t belong to the stem or root. Here we have a stem *purd. There are some variants like ipirdi or eperdi, epurdi. This stem is a very old expressive Eurasian stem *pVrd the meaning of which may also be ’fart’. For instance we shall find it in Albanian with pordhë ’fart’ as given by M. Morvan (Etymological Dictionary of Basque, online). The variant eperdi means also ’partridge’, a bird known for his moving arse (called also eperdikara with ikara ’trembling’). Although the Basque word eperdi is obviously not a borrowing from Latin perdix, it is likely that the latter goes back to the same Eurasian expressive stem *pVrd ’arse’.
Thus it is not understandable why J. A. Lakarra lets come the Basque word ipurdi ’arse’ from a Proto-Basque form *ibi-erdi the meaning of which being ’half ford, middle of the ford’, from ibi ’ford’ and erdi ’half, middle’. The way to arrive to this conclusion is overly tortuous and of course absolutely untenable: middle of the ford = way for the fecal matters = arse ! (Lakarra 2019: 392) . This is unfortunately the absurd result of the want to explain all genuine Basque words only with the help of the internal reconstruction and to refuse comparisons with other Eurasian languages.
Page 394 legen /negel ’lepra’ 1650. From Latin niger ’black’ (cp. French nigelle ’disease of plants’).
Page 399. hurran ’near’. It is an enantiosemic word. Cf. urrun.
Page 401. urrun ’far, other side’ 1448, urrin 1496 RS. for Refranes 1596. From the same enantiosemic stem *(h)urr- as in hurran, hurbil ’near’.
Page 413. iño ’to say’. Would be the stem of the conjugated forms diozu, etc. with *io. Possible if not a secondary meaning of jo.
Page 417. herots ’reputation, fame’. It means ’reputation in the country’ (cp. french ‘ça va faire du bruit dans le bourg’ ). The word is Souletin, so that the loss of one r is highly probable in this dialect, from herrots < herri-ots ’noise of the country’.
Page 418. euskara ’Basque language’. From *enuskara < enausi. Very dubious. It would come from enaus- ’to speak’, variant of erausi, erasi.The hypothesis of a derivate *ausk- > eusk- would be better. From the big Aquitanian tribe Ausci which has given its name to the city of Auch, department of Gers, the older Elimberrum (ili-berri ’new city’). The evolution a > e is common in Basque : albitz > elbitz ’hay’, arlatx > erlats ’cornice’. Moreover the putative form *enaus with fall of the intervocalic n would not give *eus but *eaus.
Page 421. josi ’sewn’ .Would come from *e-Cos-i. Dubious. The root is *os. So we remains with *e-os-i or *i-os-i.
Page 423. izeba ’aunt’. From *ize-ba. The stem is *ise.
Page 424. izter ’thigh’. From *iztar alternating with iztai like amar and amai, bizkar and bizkai.
Page 428. josi ’sewn’. From *e-os-i. The root is *os.
Page 429/430. erosi ’bought’. Would come from *e-ra-Cos-i. The root is *ros. The already very dubious hypothesis *e-ra-Cos-i does not permit to rely the verb to josi ’sewn’. There is an obviously too big semantic gap between to buy and to sew.
Page 431. ke ’smoke’ 1496 RS. for Refranes 1596. For J.A. lakarra from *gene (sic). This etymology *gene is absolutely untenable and fantastic. The word khe ’smoke’ (variants ke, eke) is a very old prehsitoric Proto-Eurasian word of expressive origin as shown by the aspirated velar or uvular kh, qh. It imitates the sound of the cough provoked by the smoke. The original meaning is both cough or smoke generally. The word can be found in Caucasic languages (Cherkess, Dargwa khe, qhe) as well as in Sino-Tibetan languages (khe, khew) and even until Amerindian languages, for instance in the Californian Hokan (cf. qhe ’smoke’, in C. Jany, Chimariko Grammar, 2009: 19). Note that the Paleosiberian Ainu has eke ’fire’.
Page 439. labur ’short’ 1072. The word is wrongly cut as *la-bur. What is *la and what id *bur ? The mistake of J.A. Lakarra consists in analyzing wrongly many bisyllabic words : za-har, za-bal, sa-bel, la-gun, etc . The word labur should be analyzed as *lab-ur. The stem is *lab- and ur is an adjectival suffix.
Page 440. lau(r) ’four’ 1174. Would come from labur ’short’. Five fingers less one. Perhaps.
Page 445. lahar ’bramble’. Curiously the word is mixed in the same group as asu and ardui ? Not possible. The link with larre ’heath’ and larri ’severe’ is dubious.
Page 452. lats ’brook’ 1013. Would come from *dats. Unsure. Especially when adats ’hair, mane’ 1571Leiz. is also analyzed as *a-dats from *dats when it is clearly a derivate from adar ’branch’.
Page 467. mihi ’tongue’, mii 1490 Garib. Atsot. for Garibay 1590. The old form *bin is still visible in the word galbin ’bell’s clapper’.
Page 479. har, haar ’worm’ 1496 RS. for Refranes 1596, ar 1520 Isasti for Isasti 1620. Does the Roncalese nasalized variant anr permit to reconstruct *anar < *nanar < from a root *nar ? Very dubious.
Page 481. nigar ’tear’ 1527 Zalgiz. for de Sauguis 1627. There is an important variant negar. It seems that J.A. Lakarra has abandoned his former absurd hypothesis from Latin lacrima > Roman (la)grima > grina > girna > nigar.
Page 484. neska ’girl’ 1496 RS. for Refranes 1596. Probably from *ne-ska < *ne- ’woman’ and diminituve suffix -ska. Variant *niska. Cp. neba ’brother of a woman’ < Proto-Basque *ne-ba. Cp. Eurasian *nV- ’woman’. Uralic *ne, Sami nis, niso(n), Finnish neis, Hungarian nő. Basque has also no for calling a woman (to for a man) and -n for the feminine of second person singular.
Page 487. odol ’blood’ 1401. The stem is *dol ’to flow’. I don’t know if the reduplication *dodol is necessary. Cf. idol, udol.
Page 488. erori ’fallen’ 1520 Isasti for Isasti 1620. The root *ror of erori is compared with the one of odol ’blood’ which is *dol ’to flow’. Rather questionable.
Page 496. orduan ’so, now’ 1490 Garib. Atsot. Wrong date for Garibay 1590.
Page 509. sabel ’belly’ From *sa-bel ? Unsure. Why should we have here bel ’black’ ? What is the first segment sa-? An analysis *sab-el would be perhaps better. The comparison with samin, samur is strange and dubious.
Page 519. senhar ’husband’ 1527 Zalgiz. for de Sauguis 1627. The stem is *sen-. The word is not cognate with Latin senior. Cf. seni, sein, sehi, seme, etc. The stem *sen- seems to qualify any member of a family (and perhaps the family itself ) and is already attested in Aqutanian.
Page 522/523. sei ’six’. For J. A. Lakarra it would come from *seni (sic) and would mean *sehi-ume ! Untenable of course. It is a Roman/Indo-European borrowing, even if there is no final -s. Cp. Italian sei ’six’. Cf. also zazpi.
Page 539. sugate, suate ’hearth’ 1496 RS. for Refranes 1596. The oldest form is sugate from 1562. From su ’fire’ and place suffix -eta, -ate.
Page 547/548. txakur ’dog’ 1496 RS. for Refranes 1596. Unproved etymology. The hypothesis *otso-kur > *(o)txa-kur ’degenerated wolf’ is too dubious. It may be also a metathesis of Hispanic cachorro ’little dog’ . Cf. the variant chacurro (1593) : el dicho baile ha tomado un galgo chacurro perruno.
Page 555 urratu ’torn’ 1490 Garib. Atsot. for Garibay 1590. Stem urra. A link with urre ’gold’ is unsure.
Page 560 uzatu ’scared away’ 1496 RS. for Refranes 1596. Var. uxatu. From an expressive uxa.
Page 562. uzta ’harvest’ 1557 Oih. Atsot. for Oihenart 1657. From *austa < *agusta < *augusta ’harvest of august’. Probably right even without the initial a-.
Page 565. zabal ’broad’ . J.A. Lakarra mentions ’no clear etymology’. From *za-bal. Why? It is not right to cut it as *za-bal. This is the reason why J.A. Lakarra cannot find it clear. What is the segment za-, what is the segment -bal ? The correct stem is *zab- / *zap- (*sab- /*sap-) with adjectival suffix -al. Cf. zapal ’flat’. From Pre-Basque *sab- / *sap- ’flat, broad’. Cf. also zahar wrongly cut as *za-har.
Page 567. zahar ’old’ *za-har. Why not *zanar (*sanar) ? Excess in continuous wrong cutting of za-. cf. also zabal < *za-bal instead of the right cut *zab-al, zap-al, etc. (*sab-, *sap-) with adjectival suffix -al.
Page 570/571. zahagi ’goatskin, wineskin’ from < *zaha-gi with gi ’matter’ for gai, gei. But what is zaha- ?
Page 581. xori ’bird’ 1527 Zalgiz. for de Sauguis 1627. The word would come from zori ’omen’. But it may be at reverse. In Latin the word auspicium ’omen’ is coming from avis spicere ’to look at the birds’.
Page 587. zurda (surda) ’hair, horsehair’ 1557 Oih. Atsot. for Oihenart 1657. Of course not from zur ’wood’ as written! The word cannot be separated from zerda 1562 Landucchio, a loanword from Spanish cerda. Probable influence of urde ’pig’ because of the hair/ silk of the pig. Spanish cerda, cerdo ’hair, pig’.
Page 587/588. zuri ’white’ . For J.A. Lakarra from zur ’wood’ ? The Basque word *sur-i ’white’ might be perhaps also compared with Turkish sur ’grey’, Chuvash sur ’white’, Samoyed siro, sira ’white’, Old Japanese siro ’white’. If cognate with zur ’wood’, it would be only possible with reference to the white birch bark.
Page 590 zuzen ’straight’ from *zu-zen. The first part is possibly zur ’wood’ (tree) like in zuti, zutik. The second part -zen would be the same as in gizen ’grease’ ? Unsure, especially if the right cut is
perhaps *giz-en. What is the semantic link between zuzen ’straight’ and gizen ’grease’ ? None.
Page 591. zoro ’crazy, fool’ 1527 Zalgiz for de Sauguis 1627. The first attested date is 1545 Dechepare. There is no reason to let come back this word once again from the stem zur ’wood’ !
Page 600. zuhur ’wise, thrifty’ < *zunur. Of course not from zun < zur ’wood’ !
Page 606. zulo / zilo ’hole’. Certainly not from zur ’wood’. The fact that there is sometimes a variant zul for zur (oezul ’bed-wood’ by Añibarro or Deen) does not allow to make of zulo ’hole’ a cognate of zur ’wood’. Unsound.
Page 609. zimel ’wrinkle’ (1366) from zumel < *zun-bel ’black wood’ ? Very dubious. It does not match semantically. Cognate with zimur ’wrinkle’.
Page 615. zidor ’path, narrow way’ 1496 RS. for Refranes 1596. The word is not very far from Spanish sendero ?
Page 616. zigor (zihor) ’little stick, whip’ from zur ’wood’ ? Cf. however Gascon sigorre ’seaside rush’ (substratic word ?) . Eventually cp. also Hungarian szigor ’rigor’ ? The latter unsure of course.
Page 617. zibo ’swing’. For J. A. Lakarra from *zi-bo with *zi- that would be an allomorph of zur ’wood’. Highly dubious. The Basque words for swing are of expressive origin. Cf. the variants zabu, gabu. They cannot come from zur ’wood’.
Page 617/618. zil ’navel’ from zur /zul and zilo ? Possibly cognate with zilo, but certainly not cognate with zur ’wood’ ! The fact that zur ’wood’ has sometimes a variant zul (oezul ’bed-wood’ by Añibarro or Deen) does not allow to make it cognate with zulo ’hole’ !
Page 618. zilbor ’navel’ < *zil-bor , zilbot < *zil-bote, zilko <*zil-ko. It is of course not possible to rely the word zil ’navel’ to zur ’wood’ as done by J.A. Lakarra ! The chain zur > zun- is correct for the trees , but not zil. There is no semantic link between navel and wood.
Page 620. zilar ’silver’. The Indo-European hypothesis (silubar, silabur and Botorrita) is mentioned. But the author cannot help to mention again the root zur (!) with allomorphs zir > zil. As he lets come also zuri ’white’ from zur ’wood’, it becomes possible to imagine a compound *zuri-har > *ziri- har ’white stone’ that would explain zilhar ’silver’. Of course untenable.
Page 622. ziri ’little stick’. From zur ’wood’ ? Perhaps.
Page 623. zirto ’point, stitch’. Probably expressive from *sirt- or eventually but very unsure from ziri-to ’little stick’.
Page 624. txirbil ’shaving, chip’. The link with zur ’wood’ might be possible for the meaning (chips of wood) but remains very dubious. The word looks expressive and means rather something light. Cf. txiribiri. Note that txiribiri means also butterfly.
Page 624. zirin ’thin rain, excrement, diarrhoea’ 1650 Pouvreau. The link with zur ’wood’ is dubious. The word means something very thin and quasi liquid. It looks expressive. Of course one could think to the sawdust. But the word has probably also something to do with irin ’flour’. Eventually a contraction of *zur-irin. Unsure.
Page 625. zirpitz ’thread, fibre, fringe’ 1745 from *zir- and again zur ’wood’ ? Dubious. The word looks expressive.
Page 626. zirtzil ’torn, dirty’ < *zir-tzil. From zur ’wood’ ? Very dubious. The word looks expressive.
Page 626. pirtzil ’wrinkled’ . Expressive like zirtzil.
Page 628. zintzarri /zinzarri ’bell’ 1527 Zalgiz for de Sauguis 1627. Root *sin(t)s-.
Page 629. zintzur ’throat’ 1620. If the word is attested since 1284 (Cinçurrçu), no need to put 1620. The stem *sin(t)s is correct. No zur ’wood’ in the word but only a nominal suffix -ur.
Page 631. zin ’faith, truth’ 1590 Garibay. Would come from zur/ zun ’wood’ because in Indo-European the word for ’true’ seems to be cognate with *dru ’solid, firm’ like a tree. It might be eventually possible for the oak (but the most solid is rather the stone or the rock). The semantic comparison with Indo-European words remains hypothetical. Unsure.
Conclusion.
The book of J. A. Lakarra et al. shows unfortunately a lot of flaws. The general form is confuse with a mixture of many different things. There are many mistakes and unrealistic, often poor hypotheses for the etymologies with often a lack of common sense and rationality negating the basic scientific principles of the general historical linguistics with self-invented or forged gratuitous own rules in order to explain all style of fantastic pseudo-etymologies. And of course we can see the very bad result of the enormous cognitive bias of the all-explained with the sole internal reconstruction seen as a holy grail for not to get hold to external comparisons with other non-Indo-European or Pre-Indo-Eurpean languages of Eurasia. It is obviously a thoroughly wrong methodology, with a blind adherence to some prejudices. This leads for instance to let a lot of words with a vague similar outside or external shape come from a unique stem with an obviously wrong analyze and cut of words like za-bal instead of *sap-, za-har from an overputative common illusory first part like *za- or like *ga(r)- from gain /gan ’top’ explaining as well words like gari ’wheat’ or even garbi ’tidy’, etc. J.A. Lakarra started unfortunately from the assumed anti-historical principle that Basque has never had another linguistic environment than the current or historically known environment. Basque is an isolate but not a complete isolate even if it cannot be included in a well-delimited traditional languages family. It has undoubtedly some connections with other old languages of Eurasia. It is not sufficient to give to one’s work a big volume and the appearance of a very scientific external shape with huge lists, all actually interesting references and big bibliographies. If the external form becomes more important than the content, the purpose remains unfulfilled, especially when the content contains unfortunately so many biases and unrealistic etymologies.
The time has come to rethink the question of the etymology of Basque, if possible without unconscious and unscientific nationalistic ulterior tendentiousness that could put some taboo on researches outside of the Basque language itself. If scholars have failed until now to find a linguistic cognacy of Basque with other Pre-Indo-European languages, in spite of their undoubtedly great erudition, the reason lies in the fact that they always tried to compare Basque each time with only one language family in a binary way. The great oldness of Basque should on the contrary lead us to compare it with the whole Pre-Indo-European Eurasia and not to try to include Basque in a precise and traditional language family. There is no other way. Basque is a Proto-Eurasian language.