Iberian and Basque (A morpho-Syntactic Comparison)

[page-n-487]
ARCH IVO DE PJ\EHISTQRIA LEVANTlNA

Vol. XVlIValentia. 198 11

W. PATIISQN
lEE. UU.)
IBERIAN AND BASQUE
(A Morpho-Syntactic Comparison )

I

When confronted with the question of Basque and Iberian rela tions, many scholars have not been content to restrain their researches to the purely linguistic field, but have gone bold ly into pre-his·
tory with theories about the relationship of the Iberian (or the Basquel
language to those of Asia, Africa, and Europe. Consequently their
ideas frequently clash.
Schuchardt (11 thought that Basque and Iberian were closely rela ·
ted; Tovar belives the two languages were related in proto-historical
times and that Iberian is fundamental1y a Hamitic tongue (2). Bot.h
Schulten (3) and Bosch Gimpera (4) believe that the Iberians came to
Spain from Africa, presumably bringing their language with them .
But J . Hubschmid (5) says emphatically that Basque and Iberian ha Y<'
nothing to do with Hamitic. Men~ndez Pidal (6) manages to find n
compromise position: the Iberians came from Africa and their lnngua

111 H. SCHUCHARDT: dHe Iberische Deklination•. Siuungsber!ehte der Kais. Akedcmie
der Wissenschaften in WieD. CLVn, 2, Viena, 1907, pp. 44 -5 and 64.
!21 A. TOVAR : u.enguas prerromanas de la Peninsule Ib6rlca•. Enc.iclopedia Lingorsllca
Hispana Ibereafter cited as ELH), I , Madrid, 1960, pp. 21 · 2 .
(31 A. SCHULTEN : cGeograna y Etnografia de la Peninsula lberieallhereafier cited as
Scbulten : Geogr.l, Madrid, 1959. vol. n , p. 61.
141 P. BOSCH GIMPERA: tlberen, Reallemon der VorgeschJchte, vol. VI. Ber lin. 1926,
p, 4 .
151 J . HUBSCHMID : .Mediterrane Substrate mit besonderer Berilckskhtigllnr. des BilS
Ir.lschen und der west-astilichen Spraeh.beziebungen.. Romanies Helvetica, 70, Bema, 1960.
161 R. MENENDEZ ProAL: 'Sobre el substrato mediterr&neol. ZRPh, 59, 1939, p, 189.

-

487 -

[page-n-488]
,

W. PATTlSON

ge was adopted by the Basques. This conflicts with Uhlenbeck's belief, namely that the Basques were not Iberianized. although their
language has some Hamitic elements cOming from the Iberian (7). Finally, Laroo declares that Iberian is not an Indoeuropean Language
(8) although Almagro thinks that the Iberians spoke an ancient form
of Celtic 191.
In our opinion, it is too early to make sweeping categorisations

such as those just quoted. Until Iberian is better known through further archaeological discoveries and more patient work in limited areas
of comparison with other languages it is futile to advance conclusions
of such a broad scope. We believe we can show possible morphological concordances between Iberian and Basque, but we also feel it
ill advised to conclude that Iberian was the source of Basque or that
Basque gave some of its structure to Iberian. We hope that our obser·
vations will be a step towards more certain knowledge . We agree
wholeheartedly with Domingo Fletcher Valls who says we can hardJy
cast aside the possibility of Basque·Iberian relationships and that we
must accept it as a working hypothesis (l0).
We shall attempt to see what structural elements in the two languages favor this hypothesis.

n
SYNTAX AND MORPHOLOGY

I For the purpose of Bnalyzing the structure of Iberian we must
firts collect all the inscriptions of undisputed authentiCity, then seg·
ment them into recognizable elements of frequent occurrence throughout the preserved corpus of the language. We can then hope to clas-

I7l c. C. UHLENBECK: cLa langue buque et la lingwstique generale•. Lingua, I, 1947.
pp. 60· l .
181 R. LAFON : d.a langue basque•. ELB, I , p. 93 , where the negative was omitted erro·
neously.
19} M. ALMAGRO: UIIstorla de EspafulI, dirlgida per R. MENENDEZ PIDAL !Hereafter
cited u Hjst. Esp.), I, 2. Madrid, 1963, p. 234: fTodo nos inclina a penlar que los pueblos ib6.
ricos del Levante y valle del Ebro hablaban diaJectos ctlticoa, es decir, indoeuropeos. aunque
de estructura muy antigua .•
1101 D. FLETCBER : lNeue iheriscbe Inachriften aus der Provint CasteU6n de la Planlll.
n le Sprache, 16. Wien, 1970. p. 168.

- 488 -

[page-n-489]
lBRRlAN AND BASQUE

J

sify these segmental elements into roots and suffixes, the latter
roughly divided into two categories - those which add a new semantic element and those which serve an inflectional or derivational purpose. To illustrate with examples from Basque : ilargi 'moon light'
compounds two roots ; ikasbide 'lesson, example' comes from ikasi
'learn' plus bide (which can stand as an independent word) 'way,
path'; ongi 'well' comes from on 'good' plus the particle - gi, ki which
has no independent use. These types of suffixes are not always clearly
differentiated, nor are suffixes always absolutely differeni from
roots. As in the case of bide, a suffix can also be a root.
2 We now look at some series of Iberian words incorporating fre·
quently occurring elements.
blur {alonel (Ill
bluibl1121
biuititeegiar {Misc. 551
sosinbiui"u (Mise. 431
Biurno (A$colil (131

blui"tetel {Misc. 34al
biuitlgl (141
blultibas (151
balcebluraies (Mise. 431

balcebiuraies (Misc. 43)
balcelagu 1161
bUosbalcarcais (Mise. 331

balceatin (MU, xxviiil
balcer (171

balset.aS (Mise. 761
belasbaiser libld. 23!
Baesadin (Livy xxxm, 44, 41
...espaiser (Ascolil
baie (alone, MU , 911

baisebilos 1181
balsegolos (Mise. 1181
aunlbalseceteban Ilbid. 441
Baesisceris (191
Tannepaeseri (eIL, Il, 58401

(J I) M. GOMEZ MORENO : cMisceltneas, bistorla·arte-erqueoJogla, Primera serib. Ma·
drid, 1949, p. 296. IHerea1ter this work is cited as Mise. and when followed by a number re·
fers to an i.nsqiption recorded in the section tSuplemento de epigrafla wm.cu, pp. 284· 330.1
(121 E. HUBNER : &Monumenta Unauaa lbericae. lhereafter cited as MLII. Berlin, 1893,
p. 15, 2 and J . UNTERMANN ; cLengue gela y lengua Ib6rica ..... APL, 12, 1969, pp. llt -2.
Biurbi appears on coin thought to be ofPerpigD8D . See U. SCBMOLL; &Die Sprachen der vorkeltischen indogermanen Hlspanlenu. Wiesheden, 1959, p. 62 . Cf. the modern place names
Biurrun INevan-al, El Plul IMedridJ. and El Plur6 del Bananc Fondo (Teruell.
1131 Tha bronze pJaque honoting Iberian soldiers, published in several places, v.g. Mise.,
pp. 233· 256.
114) D. FLETCHER: .Orieyl rn, plomo iberico e&erlto procedente de Veil d·Ux6•. A. E.
Arq. , 4O, 1967 , pp. 51·9.
115) J. MALUQUER DE MOTES y M. OUVA: eEl nuev o pJomo de UUastret.. Pyrenae, I,
1965, pp. 124 fr.
(16! From Solaig; see FLETCHER : Die Spreche (cited in note 101. pp. 150-6; elso In Ana
9, Sagunto, 1965, pp. 4-7.
1171 Mise. 9 and J . CABRE AGUILO : cCerlimica de Azailb, (hereafter cited as AzaiJel
Corpus Vasorum Hlspenorum, Madrid, 1944, fig . 21. no. 271.
(181 M. ALMAGRO : tLas inscripciones ampuritanBS sriesas, Ib6ricas y latinas-. Barcelona, 1952, pp. 68-9 .
119) Corpus Inscriptionum Latinarum. Berlin, 1885, 11,322 1.

- 489 -

"

[page-n-490]
•

W. PATI'ISON

baites (alone) (201
baldesbl (211
baitesgi (Mise. 43, twice)
•• Idutibaite (Mise. 54)

butaibaideba ISolaigl
baitiban 1221
bllos (alone) (25)

bllostibas (261
bilosbaicarcala: (Mise. 33)
ta15cubllos (27)

baldeslr (23)
baitesbaniecarM (Mise. 43)
baserokeiunba.ita (241
bald (Auna, p. 28, no. 1581
baltolo (MU, 27)

biloSfl (Villarea VI
BlIuslibu iAscoIil

bilasgre ISerreta III
baisebilos IAlmagro, Ins. Amp.,
pp. 68-69)
bostebllos lEns., Mise. p. 324)

ICOl'beles (Mise. 107)
lcortibu (Ens" Untermann. 19701 (2B)

loortai (Mise. 109)

lIdutai (Mise. 76)
ildubeleseban (ibid. (0)
nef-eildun (ibid. 45)
i1ducolte (MLI, p. 42)
bloiUdun (Serreta III
ilduniraenai (Serreta I)

IIduratin (Mise. 31)
aloii.1dui (ibid. 23)
abarlldur (MU, p . 33)
ilduro (ibid., p. 49)

Umarlllun (AscoU)

Uurbeda 'a god, B.A.E .. 'fi1t 362.
from Portugal)

Uurberri:l:o (29)

Ddum (31)

Ildunbar ... (30)

IIlurtibas fAICOU, or. Dluro,
mod. Matar6)

(20) E. CUADRADO DIAl; &El plomo con inscripci6n iberiC8 de Clgarralejo (Mule, Murcia)). Cuademos de Historia Primftiva, 5, 1950, pp. 5-42 .
(21) M. OUVA PRAT: tElnuevo plomo con inscrlpcl6n lb6rica, ball ado en UUastr"eb. Py.
renaa, 3, 1967, pp. 107-123.
(221 M. GOMEZ MORENO: cDisgres.ionas ib6rical, escritura, lenguu. BRAE, 24, 1945.
p. 276.
(23) C. VISEDO : .Unnuavo plomo escrito dela Sen-elb. A. E. Arq., 23 , 1950, pp. 211 -2
(hareafter cited as Sef1'8ta Ill ; also Ullastret, plomo m (cited in nota 21 I and Orlayl Ut tnota
141.
t241 M. GOMEZ MORENO: cOe epigraHelherice : al plomo de AJcop. Mise., pp. 219-31
thereafter cited as Serrata 11.
(26) D. FLETCHER : .Cinco inscripclones ib6ricas de LoB Villarep. APL, 15, 1978, pp.
191 ·207. See pp. 200 and 202.
(26) J , JANNORAY: lEnsenme: contribution tl'etude des clvillzatioDJ preromaines de la
Gaula meridlonale •. Paris, 1955 Ihareilftar cited as Ens.l. See plate 67, no. 29.
(271 For tals·, dais-, compare antaJsker, PLETCHER : &Nuevas inlcripclones ib6ricas de la
regi6n valencianu, APL, 13 , 1972, p. 110; arwidaJsco, MiIC., p. 245, catedalsco, ibid. and
tantindab, CIL, 12, 709.
(281 An Wlpubll5hed reading in tha MUI80 de Prehiatorie, Valencia.
(291 The name of a god in tha ValJe de Arin. See A. BELTRAN : Hlspanla Antique Epi.
graphica, Zaragoze, 1950 fr., no. 402.
(301 C. VlSEDO: «Dos nuavQl ploIDOl escritos de la Senelb. A. E. Arq., 25, 1952, pp.
123-4.

1311 J . VIVES : cInacripcionas Iatinu de la Bsp8l1a romanu. Barcelona, 2 vols .• 1971 -2,
no. 1779.
- 490 -

[page-n-491]
IBElUAH AND BASQUE

Utir (321
iltiitaialirban (33)
Utirteor (MU, 13)
iltiradm (MU, 120, 12)
auaunlltiiten (Mise. 22)

,

iltirta (MU. p. 36)
iltirc1!scen (Hill. loc. cit.)
iltirtar (34)
.. .Itirtige (35)
iltiTbidu, etc. (Fietcher, J. I .•
p. 44; Mise. 74b iIlitbite)
nwceiltira 7 wl (MU, Iv)

iltirbigisen (Misc. 42)
bastesiltirte (Mise. 65)
escer (Mise. 37cl
isearbelas (ibld, 1001
urcescer (Serrata IV)
!scer (aJone) (36)
Baeslsceris (CIL, n, 3221-6339)
Escerior IJ. Vive!, no. 56651

aiunescer (Misc. 32e)
iscaradm (Mise. p. 2671
sacariieer (Serrata I)
M . Iscer (37)
Tannegiscerrls (CIL, 37941

sacariScer (Serrata I)
lacarbetan (Mise. 411
sacarbJscar (381

saearlscer (Mise. 74g1
saearilo ... (MU, viII
sacaJ iseer (com of Castulo,
Guaden, no. 298)
saealacuca (Serrata VII (391

saearblk (CigarraJejo)
tlglnacar (Mise. 21
salduie (MLI, 351
saldulacogiar (Mogente, A, 13)
and 20 ; P. Beltr6n reeds
sa.ldulacogiabe)
!alducobaeulebobei-cur... (Ull.,
plomo rnl

saldulacogiato (Mogente. A, 141
seldugilercu IUllastret. plomo 1111

ialif (Serreta I ; Pletcher.
Cinco lnsc.1
ialirban (MU, 30b)
betaseialir (Mise. p. 278)
iltlitaiallrban (MU, 30bl

ialirg lSerrata I1

saldutiba.ite (M1sc. 54)

... nialirbiter (401
gulsceSallr (Mm, I, pp. 80 end 1531
waconSalir (MLI, 5)

1321 Serreta IV (cited m note 301; D. FLETCBER; dnscripciones iherices del Museo de
Prehlstoria de Valenciu. Valencia, 1953, p. 37.
(331 MU, p. 37 ; G. F. RILL; tNotas on tbe endent coinage of Hispania Citerion. New
York. 1931 , p. 67.
134J J . C. SBIUl.A RAFOLS; eEl poblado ib6r:1co del Castellet de Benyoleh. Ampurias W .
Barcelona, 1941 , p. 22.
(35) E. LLOBREGAT; d.os grafitos en escr:itura j6nJca e ib6rtca del esta, del Museo de
AlIcanta •. Saitabl, 15, 1967 , p. 3.
(361 J . BERGUA CAMON : .Slgnifieado y sistema de la primltiva lengua ib6ricu. Zeragoza, 1974.
(371 On a com. See A. M. DB GUADAN; diumism6tica lb6rica e ibero-romana •. Madrid,
1969, no, 169, p, 62 .
(38) P. BELTRAN : cBI plomo escrito de la Bastida de Les A1cuses (Mogenta)., Valencia ,
19&3 , second edition witb changes, 1962 ; FLBTCHBR : I . I . (see note 321. Beltrin reads sa-

carb£.sgalM.
(39) D. FLETCHBR : N. I. (see Dote 271, pp. 119-122.
(401 D. FLETCHBR: N. I., pp. 108-112.
- 491

[page-n-492]
•

W. PATTISOtl

lOIintacer {411
nabarsosln {Almagro. Ins. Amp.,
pp. 72 and 2601

soslnbluiu (Mise. 431
Sosinaden (Ascoli)

~inllSae

Soaintigi (PUny, N.B .. nt, 15J
.. .Itirtlge IUobntj8t, loc. ciLI
soenbibu tag! fA cudiel (421

are tace

{Ascolll

(MU,

vi. xxiii. xxvi)

ant legi (MU. viU

3 Our flrst conclusion from the segmentation of our tabulated
words is that Iberian did not use prefixes in word composition, which
is also true of Basque, with the important exception of the verb, in
which prefIXes are frequent. If we were to assume that a certain initial word element is a prefIx -for example, biu;' in biurtigi- another
compound -sosinbiutu- shows us that the segment can occupy medial or final positions. Similarly, baise in baisetas and belasbaiser;
and sasin in sosinbiuru and nabarsosin. These elements which appear
in either position in the compound must be radicals , the principal CODveyors of meaning.
Basque distinguishes between substantives and adjectives in forming compounds. Two substantives form a compound in which the
dtterminant precedes the dttermint : etcheljaun 'house master'; il/argi 'moon light' . If the same rule obtains in Iberian, a root which appears both in initial position and elsewhere in second place must be a
substantive. Furthermore, when such a radical (which acts as a substantive in other combinations) is preceded by another root, this latter
must also be a substantive. For example, buos is initial in bilosbalcareais and in second position in baisebilos where baise modifies or limits bUos; hence in bostebilos we take boste to be a substantive.
When Basque forms a compound of noun and adjective the order
is the opposite of what we have just seen. The substantive precedes
the adjective: etehelberri 'new House', egu/erdi 'mid day'. These adjectives are roots. They can stand alone and they can take suffIxes.
An element which we take as an adjective is beU, beZes, very common in names (Beles, Umarbeles, 5anibelser, Bennabels, all from Ascolil and in Iberian script icorbeld (Mise. 107), ildubeles libid. 40) iscerbeles (ibid. 1DOl. Beles can take a suffix: belestar (50Iaig) and belesair (ibid.). Only in one case, beleSceretorosair (Orleyl rn), is there the

1411 D. FLE'I'CHER : tTres la~idlLS funerarias !MriCILS dels VmeLS ICanet 10 Roig. Caste·
D6n).. xm Congreso Nacional de AnpIeo)opa, pp. 659·664. See p. 660.
(421 R. RAMOS FERNANDEZ: dnscripciones de la Alcudia IElcbe)J. APL. 12. 1969, pp.
168- 116. See p. 113.

- 492 -

[page-n-493]
IBERIAN AND BASQUE

,

possibility of its standing before another radical. Hence it is almost a
certainty that beu, beleS is an adjective which confirms what has
been supposed by all those who have seen it as an adjective and specifically a cognate of Basque beltz 'black',
Iberian biur has been compared to Basque bigur, bifh)ur 'torcido'
(43) but our hypothesis shows it is a noun and cannot have an adjectival meaning. Bi(h)ur and its derivatives (in Azkue 's dictionary) show a
common basic element 'twist, turn, return ; turn against, rebel' . The
form. bihurritasun 'car6cter violento. ind6cU' seems to be the most appropriate to relate to the Iberian biur in men's names, since what little we know of Iberian personal names indicates that they were often
chosen to inspire awe and fear: consider the name elements Beles
'black', Ildun 'darkness' (UmariIlun, Abarildurl. and possibly Iltir
'wolf (Iltirdesl,
In the name Nabatsosin (Almagro, Ins, Amp , 72 and'2601 the flI'St
component nabar has been equated ~ Basque 'vario, abigarrado, pardo', but as in the case of biur its place in the compound assures us that
it is a noun. (See below, p. 31).
The difficulty of establishing an absolutely certain classification
into nouns and adjectives is shown by the uses of salir, which stands
alone or in second position. Hence salir seems to be and adjective; nevertheless its appearance in the recently published Villares V (Fletcher, Cinco Inscripciones, pp. 201-208) in the combination salirbosita
throws our original notion (that salir is an adjective) into doubt. If bosita is a noun, according to our hypothesis. salir must also be a noun.
Standing alone salir precedes numerals in this document ; this, taken
with its use with city names on coins, suggests that it is a unit of value
or coinage. G6mez Moreno (Misc. p. 278) suggested a possible relationship with Basque zilar 'silver', and Tovar showed that this word
seems to have been carried to the Germanic and Balto-Slavic peoples
by the Bell Beaker migrants from Spain. Elsewhere the word for 'sUver' is based on the root arg-, (44).
There remains the puzzling phrase iunstir saliry (La Serreta I), the
first words of one side of the famous Alcoy lead tablet. If, 8S we belie-

(431 AmOD8 other, by L. MlCBEI.BNA: cComentario. en torno a la lecgua ibericu.
Zepbyru., 12, 1961 , pp. 6 · 23 (see p. 211 ; J . HUBSCHYID : ITnauna Praeromanicuat. :Pud·
kal 2, Bem, . 1966, pp. 56-59, also cMediterran8 Substrata (188 note 61, p. 73 , D. 9.
(44) A. TOVAR : lBuque Language and Indo-European Spreacb. Third IndG-European
ConferenOl, University ofPannsylvanla. 1966, Pbiledelphie, 1970, pp. 271-273.

- 493 -

[page-n-494]
•

W. PATTISON

ve, iunstii' is an honorific title (see p. 29), then salifg might be something like 'worthy' or 'esteemed', with reference to the 'worth' and
'value' of silver.
4 We can pick out from among the many suffixes that appear in
our compilations (Par. 2) certain ones which occupy a nebulous position, neither full-fledged suffix nor radical. For example, the three apparently related suffixes -ba. -ban, -bas are constantly appearing
and ODe of them, -ban, sometimes serves as 8 radical: banite (Mise.
53&), baniteor (ibid. 57) or banitewbar (Fletcher. I. 1., Lxxv). It must
have had a meaning greater than a mere inflectional or derivational
significance. We shall return to this group of SUff1X8S later.
A parallel in Basque is bide 'road. way' which as a radical can
stand alone or combine with sufflXes: bidealdi 'hike', bideburu 'crossroads'; but very frequently it is 8 suffix: eskubide '(legal) right, authority', ikasbide 'lesson, example'. In these last compounds (esku 'hand ;
right, faculty' plus bide 'way, means' ; ikasi 'learn' plus bide) the rule
of d~terminant before d~tennin~ is abandoned. Bide is no longer chief
element of the compound ; psychologically its importance is diminished and it is reduced to a suffix: although it retains its semantic value
(45). We believe that ban undergoes a corresponding change in func tion.
Another example of a word segment which vacillates between radical and suffix is take, tegi. We find take alone in three epitaphs: are
take (MLI, vi; xxiii - Misc. 45 ; and xxvi - Misc. 44). The compound
sosintaker (Canet) is also on a tombstone. A much discussed epitaph
begins with Latin.
HEIC.

are.

EST.

leg;.

SIT/ .. .
ar/ ... (MLI, vii)

in which the te of tegi is indistinct (as is the following a of ar). Two
other possibly related inscriptions are soanbidu taki (Alcudia, iv, on a
pottery shard) and the fragmentary sbogUace (46). It seems self-evident that take, tegi means 'place'. Sasin, often attested as a man's na:
me, subordinated to or modifying take, would give 'Sosin's place ' or

1451 In old Basque compounds, independent w ord! can become suffixes and suffixes can
become independent w ords. See L. MICHELENA : BSL, 53, 1958, p. 230 and C. C. UHLENBEC1t; Lingua, I , 1948, p. 69.
146) D. FLETCK!R : N. I., p. 10 8. The inscription may be on a tombstone.
-

494 -

[page-n-495]
IBERIAN AND BASQUE

9

'Sosin's grave '. Here take is a radical and perfoms as such in the compound (471.
What seems to be the same element is found in biurtigi (OrIeyl In)
and in the many place names preserved by classical authors: Artigi ,
Astigi, Cantigi, Lastigi, Olontigi, Sosintigi and Saltigi (MLI, Prolegomena. par. 37). Here the status ofradica) begins to give way to that of
suffix. We are of course reminded of such Basque formations as ardandegi ' wine shop' , lorategi 'flower garden ', apeztegi 'priest's house', and the abbreviations of this element (- gi, -ti, -di, -dui, and doil which are clearly suffixes and which , we believe. relate to Iberian -ti, - gi in combinations with -bas: boutintibaS (Misc. 261, tasbarigibaS (ibid. 32c), and rarely with -ban: aScertiban (ibid. 75).
We shall return to these suffIxes later, when seeking their meaning.
5 A large number of sufftxes must have had derivational or inOectional values. Some appear as a single sign : sosinbiur/u (Mise. 43).
alorilduli (ibid. 23). abarildulr (MLI, 221 ; others contain two signs:
biur/bi (MU, 21 baides/bi (Ull. plomo 31, ustalailbi (Mise. 74c) and eban
en (MU, 47, 48); elsewhere we are in doubt as to whether the sufflX
is -r or -er: taker (see above), esc/er (Misc. 37cl. benebetanler libid .
671, baldusler !ibid. 119 = Liria Lxxvi). Other combinations are - iu:
ecariu (Misc. 431. aduniu (ibid.), cosoiu (ibid.) and borberoniu fibid.l ,
which all coming from the same document suggest some sort of gram matical agreement; -te in baser/te (Mise. 53a), caresir/te Iibid.). and
bante (ibid. 68) which recalls the suffix -ite : ban/ite {Mise. 53al, caresbanite (ibid. 54). For a final example of a brief sufflx we take - ik.
so written in Ionian script: sabarik (Cigarralejo), legusegik (Serreta I)
which had to be written -ice in Iberian, since this writing had DO sign
for c (or k) standing alone : Abartarice (Misc. 74dJ.
6 Some longer suffIXes which will demand our attention are
-sken which always appears on coins in co~unction with city names;
- egiar, also found as an independent word but always, we think, in
close relation to a preceding name; and lastly the pair -tar, -ar,
usually thought to be identical in ·value.

147) The same root gives LatiD tego, teguium, and toga, as well as Celtic tegi4 ·bouse· IA.
HOLDER; ~ltceltilCher Sprachschaw. LeipziJ, 1891 -1913 , cols. 1788·91. Most scholars reluctantly aC(;ept Celtic as the source of Basque tegi IJ. HUBSCIIMID : op at. in note 14, p.
I II J. The root originally bed the Idea of 'coverinS' but in Basque it lost this notion 8 S an essential part of Its meanins and kept the idea of an enclosed or limited space. Basque also has
the word told 'place'.
L. MICHELENA : IComantariO .. .I . p. 15, a ccepts the interpretation of Iberian are take as
S
·thls is the place'.
-

495 -

[page-n-496]
10

W. PATTlSON

This survey of the Iberian roots, their compounds, and the types
of suffixes shows a defmite similarity to the structure of Basque. Both
languages are to a large extent agglutinative, compounding elements
which also exist independently, but they are at the same time inflectional, using suffixes which have no independent existence. We now
turn to more specific comparisons, principally of morphemic function, but also of the semantic value of certain sumxes.
7 We begin with -en, the subject of an important study by
Men~ndez Pidal (481. He fmds this pTe-Latin element is pecu1iarly
characteristic of the Iberian region of Spain. «La mayor abundancia
de - en corresponde alas actuales provincias de Valencia, Zaragoza.
Ruesea y Lerida, es ciecii, al terrltorio mas 'puramente iberico ... » (po
6), It can form part of personal names, such as Suisetarten, Sosinaden, and Nalbeaden (ibid" all AscoH names) , In toponyms it is also
abundant in the East and South, while scarcely recorded in the North
and West, Typically, such a toponym in based on a personal name ,
hence place is described as t(de fulano», exactly as Basque forms Mi chelena cMichel's place», literally «the .of Michel» (pp. 32-34l.
In Iberian inscriptions we fmd a noun iltirten (Misc. 22)' iltirbigisen (ibid. 421. antorbanen (ibid. 69l.lacugiecen (49); urtisen (50)' aboten (ibid.) and several examples of ebanen (Misc. 15,47, 48, and 761.
EnserWle offers us ibesoaen (PI. 64, 23), but the residents of this site
preferred the other signs of possession which we now examine.
Menlmdez Pidal points out that - en often alternates with - an(p.
71 so that the same place name may appear with either ending. Ibe rian inscriptions reveal sesgersduran (Serreta I), bitan (Solaig), uwisanwi (Ens., PI. 64, 27) (51), and auetii-isanwi (ibid. , Pl. 64, J IJ.
8 Another grafitto from Enserune (PI. 64·, ) 2) is almost identical
with the last one cited from that town. It reads auetifisarwi and com pels the conclusion that -ar has the same function as -an. Now - ar
has long been lumped with -tar and taken 8S an ethnika, because in

148) R. MENENDEZ PIDAL: eEl sufijo -en: tU difuslon en la onomAstica hispana.. Eml!rita, 8, 1940, pp. 1-36; for examples from Valencian toponyms see M. SANCHIS GUARNER:
dntroduccion a la historia lingulstica de Valencia.t, p. 61 : and R. MENENDEZ FlDAL: .,-oponimia medlterrAnea y toponimia valenciana primitiva•. Bol. de Dialectolog!a Esp.. 33, 1954.
149) D. FLETCHER: tUn bronce escrito del poblado ib6rico de San Antonio IBecb{, Caste·
lion I•. Zephyrus, 1S, 1967, pp. 79·83. See p. 80.
(50) R. LAFON: dnscriptions en caracteres iberes de Perpignaru. Rev. Internat. d'ODomastique, 17, 1965, pp. 1-6.
151) The ending - wi is a frequent final suffix and seems to mean '1 (amI'. These Ensilrune graffiti on pots would indicate ownership: I am of uwis, etc. TOVAR: ELH: FLETCHER:
.
Die Sprache, 16, p. 158, n. 19: and L. MICHELENA : .Comentarios ...• , p. 19.
_ 496 -

[page-n-497]
11

IBERIAN AND lIASOUE

Basque these two suffixes are so used: Bilbaotar, Irunar. There is no
reason to jump to this conclusion, as we shall see later. The -ar suffix
has recently been seen as a possessive (52), which would of course
still permit it to indicate origin 'of or from Madrid' as well as 'belonging to Madrid',
Enserune offers other graffitti in -ar: unticorosarwi (PI. 69, 20),
selgiterar (PI. 64, 19) and abargitibcciar (PI. 64, 6) ; the Ascoli bronze
gives the names Luspanar and: Arbiscar; G6mez Moreno records dursaurar (Mise. 7), egosonar (ibid. 32b), minar (ibid. 19), enatilar (ibid.
118), besides the abundant egiar which we shall take as a different
morpheme. Elsewhere -ar appears in dusgitar (Orleyl nIl and carcoscar (Solaig). While not denying the possibility of another identical
sufflx -ar with a different meaning, we take it to be a possessive in
many cases. We shall return to this suffix when we study the combination -aren.
The -en, -an morpheme recalls the Basque particle -en, - n
which makes the word or phrase to which it is affixed modify that
which follows (53). Examples: Michellen/a 'the (or that! of Michel';
aur/du/n/a 'the (woman) with (or possessing a) child' (literally 'child she has it- particle to indicate modifying of the following - that one'
equals 'the one that has a child'). Traditional Basque grammars call
this particle a genitive which makes it difficult for them to explain its
use with verbs and clauses (54). Traditionally, a second genitive, -ko,
is recognized, and we see it also as a particle indicating modification
of the following word.
9 Iberian has an infix -cu-, --co- suspiciously like the Basque
-ko-. The words besumin/cu/egiar (Misc. 55), duSeratine/cu/egiar
(ibid, 56), and ars/agis/cu/egiar (55) all end in egier which we hope to
show has a meaning something like 'lord' or 'master' or 'magistrate '.
This would allow us to interpret the words given above as 'the lord of
besumin', the lord of duseratine ' and 'the magistrate of the Arse agis'.

(521 A. TOVAR: cLu monedu laguntinas y otras notal soMe iIlIcrlpcionel lbericas..
BSEAA, IS, 1949, pp. 25· 34. G. BARB: cBukiach und lherilchJ. Eusko-Jatintza, 2 , 1948, pp.
1- 119. See p. 106.
(53) In our view It 11 axactly like the Chlnese d. in function.
(54) A. TOVAR: .Una expUcad6n del lufljO VIUCO -en.. Primitlvas llnguu ltispwcas,
pp. 90· 95 ; 11.. LAFON : cLes delll: gemtifl du Basqut». BSL, 61 , 1965, pp. 131 - 159.
155) MLI, 4Ob; A. VlVES : I, 35 ; J . UNTERMANN : cMonument8. Linguarum HispanIca·
rum.t. Wiesbaden, 1975, 1,232 ; and M. C. PEREZ ALCORTA : .Monedas antigua.s de Slgun·
to_. Numario HlspAmco, 4, 1955, p. 278.

-

497 -

"
•

[page-n-498]
"

W . PA'I"I'ISON

In this last example, found on"8 coin of Arse (Sagunto), which coins
often bore magistrates' names, it could be that agis (561 signifies
'mint' .
Tovar (57) segments this same coin legend arsa - gis - cu - egiar
saying that gis may be related to the gita of other coins, and, most importantly, «En cu podrfa estar algo asi coma el ko del genitiv~ vasco ....
The particle -cu- is not confmed to use with -egiar. Tals/cu/bUos
lEns " Pl. 68 , 25) and ndu/co/ite Ion a coin, MU, p , 42; A, Vives, 1,16)
contain very common roots bilas and ildu. (The latter appears in several toponyms, such 8S ndum, a stage on the Roman road north of Sagunto.) Elsewhere we fInd sagustico or saguntico (58), castilco (59),
abilico (MLI, xiv), balcelacu (Solaig), and the names Austinco (Ascoli)
and Urcico (elL, n, nr. 2818). Michelena (GO) gives examples of aothroponyms in medieval Basque ending in ~o: Larraineco 'de la era',
BaztelTeco 'del linde, margen' , etc. They show that the person in
question was from a certain place, generally not a town. The -co im·
plies modification of an understood, vaguely conceived, noun or: pronoun, such as 'person', 'that (one)' , or 'he', just as in the Iberian forms
listed above . With these Basque names we can also compare the Aquitainian Estenconis (Cn, xiii, 271) which Lafon (51) calls mom d'homme Asuffixe -co, flllchi comme un nom latin en - 0 , -onisl. The leader
of the Editani called Edeco or Edesco (Livy, XXVJll, 17, 11 has a name
formed by use of t.b:e same suffix:
The existence of such pairs as eueko andere and et!ekandere
'ama de casa' implies that the Basque -ko can be reduced to the consonant k, especially before a following vowel. Another example: su-

(56) Latin ago, Greek.o/)'c.l had once the sense of 'to weigh'. The derivative agina 'a babeam' (ERNOtIT et M'EILLBT, Dict. Etym .. pp. 16·7) recallslibra ' the beam of a scalas'
hence the 'scales' and 'pound'. Obviously weighing out the matals was a prime function of
the mint. We must remember that Iberian coinage began under the Greek and (principally)
Roman influence, hence the comparison of agCs to ago fa justified. However, GOMEZ MORENO thinb this coin legend should be COITected to arsbigCsteegiar (M.lac., p. 278) on the ana·
logy to Utirbigisen (Misc., 421. Note that this is tha only Arse coin with the IStronp r.
(571 A. TOVAR : op. ciL in note 52, p. 29.
(58) FLlITCHER: I. I., p. 29 ; TOVAR: PriDrlUvallanguu hlspinlcas, p. 212, note on p.
'3,
1591 M. ALMAGRO : ZephyruJ, 2. Salamanca, 1951 , p. 104 and Ins. amp., p. 260.
(601 L. MICHELENA: cHow lingi1fst1cu a la colacd6n diplom6tica de Irach.,.. Fontes
Unguae Vaaconum, I, 1961, p. 40.
(6I) R. LArOH: &Sur la langue des aqu!tainI et celle das VUCOOD. Dul. Philologique et
Hiltoriqua, 1958, p. 3.
I ~ce

- 498 -

[page-n-499]
IlIERIAH AND BASQUE

13

kalde 'fire-side' . But is the tin sutondo 'flIe-side' simply a variant of k
and ultimately a derivative of -ko-? Basque scholars do not speak of
this possibility, although they give examples of the use of the infix
-t-: atsotiz 'proverb' from atso 'old woman' and itz 'word', ilartiz
'epitaph' from ilar(ri) 'tomhstone' end itz 162).
Looking now at Iberian, city names like lliItlurgi and Arce/d1urgi
show the same construction. Untermann (63) believes that ili - , ilu-,
in toponyms comes from Iberian ilti-. iltu-, and in his map number 2
he shows the distribution of cities with this name element. They fall
in the narrowly defined Iberian area, along the Mediterranean coast
and in the Tartessian area of Andalucia. Not all of them have the - tinfIX.

A form like biur/t/egiar shows the same inflX.
10 The suffix - ite reminds us of the names of Iberian tribes as recorded by the Greeks and Romans: Ceretes, Indigetes, Ilergetes, and
so on. This ending. to which the Romans added their sign of the plural
(-s) , just as the Greeks added -oi, was then extended with the adjectival -anus to form etanus, -itanus, to our mind a compound sufflX
(64), but considered by some great authorities as part of the W~stern
Mediterranean substratum (65). Unfortunately, examples of the tribal
names recorded by ancient authors are not found in Iberian script.
We find -ite and -te frequently in our documents and while there
is no assurance that they are the same morpheme, or even that they
always have the same function, there are reasons to assume that at
least on occasion they serve as a plural sign. They precede the abovementioned -egiar in several cases: caresirte egiar (Misc. 53a), caresbanite egiar (ibid. 54), ebifteegiar (ibid. 53b) and biuttiteefgiar] (ibid.
55) (66). If egiar can be taken as 'lord', 'magistrate', these phrases
would stand for 'the lord of the caresir people' , and so forth, the -te,
- ite indicating the plural.
We have already spoken of -ar as 8 possessive and its confusion
with etar, -tar on the basis of Basque analogies. We incline to believe

(62) R. LAFON : cS\lr un Iuffl.z:e nominal commun au bisque et quelques langues c lucuI·
queu. BSL, 44, 1948, p. 144.
(63) J . UNTERMANN : cRstudio IObre 1u areas llngQlIticu prerromanas de la Peninsula
Iberlcu. APL, 10, 1963, p. 173.
(64) A. GRAUR : cDouble luffuxation des nolU del habltant». Acta Antiqua, Budapest,
10, 1962, pp. 119-121. makes I convincing argument for the compound suffu:.
(65) J . HUBSCHMID : lMaditerrane Substratet, pp. 71 -72; POJCOIlNY: ELH, I, p. 6.
(66) Since this word tl preceded and followed by words ending in -egiar the reconstruction by GOMEZ MORRNO 11 almOlllt certainly correct.
- 499 -

[page-n-500]
W. PAT1'lSON

14

that Iberian tar, -et4T includes the plurpl, Coins with legends like Arseetar IMLI, 40) and Saitabietar lA. Vives, LIim. vi, 181 parallel
exactly the Greek letter inscriptions on Spanish coins which use the
genitive plural. 85 in EMnOPlnN(A. Vives, II, 191. Arseetar means 'of
the people of Arse' just 85 the Greek signifes 'of the people of Empo(S7).
It would be convincing if we could prove that -tar is never a sin-

riOD'

gular and -ar never a plural. The present state of our knowledge does
not permit such 8 proof, but it does not present us with a categorical
refutation. A pot with balacertar (Mise. 38c) could mean that it belonged to a family named .Balacer; Ur9idar (an Ascoli name) can be a
'man of the Urgi people'. The graffiti on Enserune pots ending in -ar
(see abovel seem to indicate possession by one individual. as does dursaurar (Mise. 71 on an Ampurias vase (68) .
Now Basque has a plural in -eta (ameta 'stones') which many
have thought derived from Latin -eta, the plural of the collective ending -etum (69). But the antiquity of this plural in Basque seems evident when we fmd traces of it in the inflectional system of verbs (70) .
The verb incorporates the object; du 'he has it' contrasts with ditu
'he has them'. and zuen 'he used to have it' with zuten 'he used to have them' . The third person plural ends in -te: dituzte, dute, etc. R. Lafon (71), points out that de suffixe -tzu figure dans des toponymes ou
il exprime la pluralitk, l'abondance. C'est sans doute le m~me sufIlxe
que sert, sous les formes -zu et -tzu suivant les parlers, a former des
adjectifs et des substantifs exprimant l'abondance ... La semi-occlusive -tz sert aussi en basque ... a indiquer dans les verbes le plural du
sujet ou dci patient ... It seems that B~sque and Iberian share - ite/ - te
as a plural sign.
11 Quite a few Iberian documents have numbers written with
vertical strokes, just as the Romans wrote one, two and three (72).
(671 U. SCKMOLL: cDie Spracben der vorkelti!:chen lndogermanen Hispaniens ... I, p. 70,
suggests that the suffix -6tar has something to do with the tribal names In -etes recorded by
ancient authors.
(68) A. TOVAR: BSEAA, 16, 1949, p, 26, maintains correctly that Basque -6tar ls lndefi·
rule in respect to number, and implies without proof that tbJs Is also true of Iberian.
1691 SCHUCHARDT : cDie lherIscbe Deklinationt, p. 24 and ZRPh, 30, p, 6 fr.; R. LAFON :
BSL, 55, 1960, p. 196; MICHELENA : Archivum. 8, 1958, p. 46, n. 27 is unsure about the La·
tin origin of -eta..
(701 L. MICHELENA: loc. dt In. 691 . ... DO parece que no pueda senalar en Jas formas
personales del verba vasco ningWl afijo que sea de orlgen latinol.
1 11 R. LAFON : d.e verbe baaquB auXVJtl sl,cl.,., I, p. 530.
7
1721 A few examples : Aulla, fig . 16, Dr. 4a; fig . 17, Dr. 30b; and fig, 18, nr, 4 ; Serrata
VI; Orley} m ; Ens. PI. 65, 21 ; Liria, plomo I (lILETCHER: I, I .. p. 42 1.
- 600 -

[page-n-501]
IBERIAN AND BASQUE

"

This same vertical stroke is also a sign of the Iberian syllabary, -ba,
and although almost all Iberian signs are clearly related to Phoenecian and Greek symbols, this one has no certain early prototype (73).
We feel that the Iberians, accustomed to associate the vertical stroke
with 'one' , called it ba because ba (or a word emphasizing this syllable) meant 'one' in their language. The analogy to Basque bat 'one' is
evident, and we note that bat always follows the word it modifies.
contrary to the usual preceding position of other numerals.
Now if ba can be equated to 'one' , it does not follow that it always
has a numerical significance. It could become a morpheme or functor
word, similar to the English one in combinations like the small one,
the saver one, the Florida one.
A number of place names in -ba, -uba, -oba come to mind: Salduba, Uduba, Corduba, Onuba, Maenuba and Iponuba (74). With
them we place Ilipa (Strabo, rn, 2, 21 Astopa (Livy, xxviii, 221 and
Onoba and Ossonoba (Strabo, rn, 2, 5). If the roots of these words have meaning as common nouns or adjectives, which we think probable, then the suffix changes the root into a place name. IfSalduba (751
can be related to Basque saldu 'sell' , the root plus the suffix must be
'(the) selling one' or 'the market place'. It can be compared to Saltigi
(MLI, Prolegomena, 17) which we also interpret as '(the) place of Salt
(or SaidI', and, if related to Basque saldu, 'the market place'. Saldui·
til ba/ite (Misc. 54) may h6.ve something in common with Salduba, as
we see when we segment in remembering ban/ite (Misc. 53, 54, 57
and 61) and such personal names as Boutin/ti/bas (Misc. 26) and Bilusltifbas (Ascolil. nlurltilbas (ibid.l.
The use of -ba as a formative of place names is not necessarily
the only way the SuffIX was employed, but since we do not know the
meaning of the radicals to which it is afflxed in the following examples. we cannot be more specific in our analysis. We find culesba (76)

(73) J . UNTERMANN : Em6rita. 30. 1962, p. 288. But J .. M. SOLA-SOLE: d4isctlllnea
p6nico-hi!panlLt. Sefarad, XVI. 1956, p. 339, believes the Iberian I (ba) comes from Phoene·
dan 1 (bet). In bb table I , opposite p. 334, he traces the evolution of the Iberian sign. Howe·
ver. he does not find bet 85 I before c. 180 B. C., which could mean thet the Iberian script in·
Ouenced the Bilpano·Punic and not the revllll'S8 . M . ALMAGRO GORBEA : Rev. de la U. Com·
plutense de Madrid, 25, 1976, p. 52 finds I (ba) u early as 600 B. C.
(74) A. SCBULTEN: Geogr .• U, p. 37.
175) Salduba, also Salduie, the name of Zaragou.,
(76) Compare culeiurla (Mise. 20).

- 601 -

[page-n-502]
"

W. PA'l"l'ISOH

lEns. PI. 65. no. 251. balacoiaba IMU. xixl. bobaitinba IVillares VI.
ba/maceba IHAE. no. 5351. bastaibaitieba 177IISolaig ~and [slosintiba
(Fletcher, Bronce). This last one will receive our attention later.
Finally 8 pottery fragment from La Alcudia (Elche) is inscribed
balcatica eba, which brings to mind the many coins heraing eba (as
well as etaban and ban) (78) which some have taken as marks of
value 1791.
The suffixes -ban and -baS are abundant as the final syllable of
Iberian words. We have noticed ntirdasalirban on 8 coin of Lerids.
On tombstones we frequently find -ban, -eban, and -ebanen, for
baisetasexample igoiwceiwi ildubele~eban (Mise. 40, see p. 322) and "
ilutaSeba[neJnwi (Mise. 76). Now although Tovar stoutly maintains
that eban means 'stone' (80). Cara Baroja and J . Vallejo both take ban
to be the equivalent of Spanish de (81), G6mez Moreno hypothesizes a
relationship between ban and the Semitic ben and .G. Bahr equates
ban with fuius (82). Following these last opinions and remembering
that wi is widely accepted as 'I (am)', the first inscription r~ads 'I
(am)lgoiwcei, son of ndubeles' and the second is '1 (am) Baisetas, son
of Iluta~· .
In other cases -ban, -eban cannot be translated by 'son' although the possessive is appropriate. Just as we postulated a meaning
'one' and a morphological function for -ba, so we regard - ban andhas as the same kind of morpheme. IltifdaSalirban on a coin means
'(the) IltifdaSalir one' as opposed to the ones of other cities. A stone
weight topped by an iron ring is inscribed ustai(n)abararban (Almagro, Insc. amp., pp. 78-80, Misc. 141. Since Basque ustai means 'ring,
hoop' and abar often forms part of men's names, itis tempting to interpret the phrase by 'ring (the I Abarar one', that is, 'Abarar's ring' ,

(77) Compare baiil (Azaila, CVH, fig. 19, no. 158), and ... tibaite (FLETCHER, I. I .. no.
XXVII).
(78) MU, pp. 3D, 33, 43, and 48 ; G. R. HILL: d.nclent Coinagel, pp. 67-8, 102, and 131 .
(79) A. VIVES: n, part 2, p. 10 ; GUADAN: &Numism'tica lb6ricu, p. 161. But since ban

is not confined to one coin value it can hardly Indicate value. For example we find it on the
reverse of semis, quadrans, and sext.ans of the mint identified by the magistrate Aharildur
(HILL, pp. 102 and 131 ; MLl, p. 33).
(80) Lex., pp. 305-6 ; Ill! euskera, sus parienteu. Madrid, 1959, p. 45 ; Prim. Leng.
Hisp., pp. 61 ·66. But is eban always written on !tone? R. RAMOS FERNANDEZ : APt: xn,
1969, p. 172, saVl it is not; an nample is Ens., 'DI. 70, nrs. 9 and 11 , where reading ~
right to lea we have ebanbonar stamped OD cla, .
18l) J . CARO BAROJA: ISobre el vocabularlo de 10 lnscripclonel ib6riclW. BRAE, 25,
1946, p. 183' J . VALLEJO: «En tomo. UDa vleja moned. iberiCat. Emllrita, 14, 1946, p_250.
(82) M. COMEZ MORENO : Mise., p. 280; G. BAD : p. 420.
- 602 -

[page-n-503]
IBE:RIAN AND BASQUE

17

where -ban has the same possessive role that we have seen elsewhe·
re (83). 'Baisetas (the) Ilutas one' (Misc. 761 indicates a Baisetas diffe·
rentiated from others of the same name as the one related to Ilutas. In
the same wa.y Lydia.n nanna.s bikivalis is literally 'Nanna the one of
Bikiva' {B41,

Most epitaphs of the type «Xl ban» seem to refer to X - a man son of Z. But there is no reason why X cannot be a woman, the wife or
daughter of Z. From Liria we have a fragment inscribed ... ban unseel·
tegiar ban (Fletcher, 1. 1., no. 1; see ibid., no. LXXV and Mise. 57) over
the picture of a lady with a mirror. If our interpretation of egiar as
'lord, magistrate' (see below, p. 29) is correct, unsceltegiar must be a
man and the phrase - if it refers to the lady - must mean 'the one (i.
e. wife or daughter) of unsceltegiar'.
Some other examples of -ban are balcebereiwbarte balduSer ban
(Misc. 119), baleeatin isbetarticer ebanen (ibid. 47), and alordui be·
laSbaisereban (ibid. 23). These names parallel those of the Ascoli
bronze in structure, where each of the honored Iberian. warnors' na·
mes is followed by this father' s: Sanibelser - Adingibas)riliu~ , Dlurti·
bas - Bilustibas f. , and so forth. The father' s name has ho genitive in·
flection, so qilius] serves the same purpose as the Ib.e.r ian - ban or
-eban. Functionally the Latin and the Iberian are the same, but - ban
cannot mean 'son' as this meaning could not translate nW"dasalirban
and similar inscriptions.
The suffix -bas shows up in the Ascoli list not only in the three"
names just mentioned but also in Umargihas and Luspangib[aSj (85).
In Iberian script we have DasbarigibaS (Misc. 32c), BoutintibaS (ibid.
26), AlostibaS or BilostibaS (Ens., PI. 67, 29), Baceocebws (Alcudia) on
a potter's stamp, and Urcebas (Mise. 53a; CB, p. 757) next to the head
of a mounted warrior on a vase painting from Liria (8G).

(831 Abar is also Celtic (HOLDER, I , col. 61. It seems to be a cognate of the Latin avlU'US,
fundamentally 'e~, dellroUl' (ERNOUT et MElI.I.ET: cDlctionnaire etymologiquet, p. 55).
In AquitanJa we flnd three place names derived from this anthroponym : Avaray, Averan,
and Averon lA. DAUZAT et CB. ROSTAING: eDict, etym. des noms de ijeu de France.• Paril,
1963, p. 411While Basque abGr 'bruch lof tree)' could bardly be cognate with the Iberian, still B81'
que abe ' tree' also means 'column, IUpport.· Abe comhlDed with the suffix - could give 8 name meaning ·support'.
(84) J . FRlEDRICR: Extinct Languages (translated by P . Gaynor), London, 1962, p. 111 ;
A. HENBECI:: Lydiaka, pp. 68-70.
(85) This name wu too lon8 to fit in the space allowed for it. GOMEZ MORENO : Mise.,
p. 250, completes it.
(86) These last two names are the only ones with the Iweab s in -bas.

- 503 -

[page-n-504]
W. PATTlSON

18

Urci, Urgi is abundant in Iberian place names. A town of Urci in
southeastern Spain issued coins with southern Iberian (sometimes CBlled «Tartessiam) inscriptions (MU. 116 and CB p. 708). It is mentioned by Mel. (IT , 94) and Pliny (N.H., rn, 10). Pliny also records lliturgi and Isturgi (CB, p. 795); and an Iberian coin bears Arcedurgi (MU
23). Among personal names we note Urgidar (Ascoli), Ureeseer (Serreta IV) and the Latinized Urcest8r Tascaseceris f. (Cl!, IT, 2067). We

suspect that Urcebais is a proper name based on a toponym and that
-bas is comparable to -ban. Hence Urcebas is '(the) Urci one' or 'the
man from Urei' or perhaps 'the ruler of Urci'. Care Baroja calls bas
run sufijo 6tWco) (CB, 714) and says that it «indice procedencia y origem (ibid. 792). His thought supports our hypothesis which differs
only in relating - bas to -ban and finding their basic meaning in 'one'
used as 8 functional morpheme.
In Basque bat 'one' (with obvious similarity to the Iberian) is an
unusual word in several ways . As mentioned above. it always follows
the word it modifies. whereas other numbers regularly precede . Its fi nal t is strange as this letter is very rare in final position (87).
In compounds bat shows we are dealing with numbers, for eXBIDplezeinbat. zenbat 'how much, how many', onenbat ' as much as this',
bana 'one apiece ~ and bakar 'one alone, unique '.
'Oneness' involves the notion of 'completeness' , of bringing together all the parts to m~e one, as in 'unite'. In Basque this idea pervades batzar 'congress ', batze 'harvest' and the suffIX -bete as in eslcubete 'handful', and betekada 'satiety, bellyful' .
None of these or the many other derivatives of bat is an exact pa rallel to the Iberian -ban, - bas. Howerer, we do find other precise
counterparts. The family name Larrebat is literally '(thel pasture
one(s)' (88). Even more interesting are the relational terms ending in
- ba. -pa, such as aizpa 'sister' of a woman, arreba 'sister' of a man,
ugazaba 'employer, master' , giharreba 'father-in-law or mother-inlaw' , and osaba 'uncle', to name a few . The suffix on these words has
defied the efforts of the etymologists (89).
Nonetheless Joan Corominas comes very close to the solution we
shall propose when he shows that aizpa is related to aiza 'compaii.ia'

(8 7 ) Bat may come from en earlier "b
(88) MICHBUNA : op. cU., p. 30.
(89) The diJtinguished Buque scholar LIDS MICHELENA says Qli int.encl6s no podrla
SBr la de penetrar B1 sentido nunce revelado y a cuo irrevelable del sufljo - beD I_sabre a1gunOl nombres YUCOl
in Foetal Unguas Vasconum, I , 1969, p. 120). But BAHR,
p. 422, flnds - ba re
nt of Iberian -ilban.

C:&=tesCOJ

-

604 -

[page-n-505]
mlUUAN AND BASOUE

19

and must mean 'compadera' (90) , Carrying his investigation farther
we note that other compounds ·of aiza imply work in common or common interests: aizaberro 'rotura hecha en comun por 105 vecinos', aizalan 'trabajo vecinal', and aizetJe 'casa de la villa' , It is clear that
the idea of sharing, especially of work, is fundamental in aiza, hence
aizpa is 'the work sharing one ', and the sutllx is, as in Iberian, 'one'
used as a morpheme ,
The term ugazaba goes back t.Q ugatz 'breast' , which by extension comes to mean 'mother's milk', and according to some 'food ' in
general (91), Words which designated stepfather and stepmother ugazaita, ugazama - are undoubtedly based on the notion of feeding
an orphaned child, In like fashion the young apprentice or. farm laborer lived with and was fed by his ugazaba 'the feeding one', where
the SuffIX -ba is the same morpheme as in aizpa.
Finally, giharreba, based on giarre 'rencor, amargo recuerdo;
parte magra de la came' seems to mean 'the disagreeable one' , and
osaba may have something to do with osabide, osagarri 'cure, remedy' for ill health, and osagUle, osalari 'doctor'. An uncle may have
been 'the curing one ' in times past, perhaps not so much from ill
health as from other difficulties. The special relationship between an
uncle - especially a maternal uncle - and his nephew has been observed by many antbropoligists. For example, Claude Levi-Strauss finds
that a severe father's son tends to gravitate towards his mother's

brother 1921.
12 Can the final letters of -ban and -btti be explained? There can
be no masculine/feminine distinction as both endings can refer to
men, But a close examination of the epigraphical evidence shows that
-baS never accompanies wi 'I (am)', On the other hand the potter's
stamp Beceocebas names a living, active man (931.
Basque contrasts the agiens, the «subject» of a transitive verb with
the patiens, the «objetct» of such 8 verb or the «subject;) of an instran-

'901 J . COROMINAS: 1T6p1ca hesp6riCD, Madrid, 1972 , p, 322 .
(911 R. M. DE AZI:UR gives these meanina, but MlCHBLENA: op. dt., note 89, p . 123, in.
lilts that ugatJ means only 'brflll8t', He accepts, however, the translatIona of ugatzaita 'pa.
dre Dutriclo', ugatzama 'madre DUtricio', where the Idea of food is essential.
AZJWE takes tlgcuaba from ugau plus aba. 'padre'; MICB:B.LBNA derlveslt from tlgal
aDd asaba 'u08StDr'. It is difficult to see how U 'antepuado Dutriclo' 'MICHELENA's own
words, p. 1241 could be 8 living muter or employer.
(921 C. LEVI·STRAU5S : cL'analyse structurale an l.I.nguistique at aD uthropolO
gie».
Word, I , 1945, p,
(93) G. BAHR, pp. 440-«1 , su.rm.lses that tha n a.nd the, of haft and bas are fnflectiODal,
, beins a Dominative aDding.

«,

.

- 505 -

[page-n-506]
20

W. PATI'lSON

sitive or a copulative verb (94). While there is no phonological similarity between the Iberian and Basque inflections, we do suggest with
hesitation that the grammatical categories are the same: ·i.e., that
-baS is the agiens and -ban the patiens.
In support of our suggestion concerning the ending -5, we turn to
A. Martinet (95) who says the ending -s of the nominative in lodo-European languages was once «proper to all nouns designating entities
capable of being conceived as agents», but not applicable to «patients».
It «was necessarily the morpheme of an ergative case, designating the
agent of the action. This was not the nominative ... An ergative case is
an.indicator offuDCtiOD ... An evolution in the structure ofl. E. had the
result of making the former ergative undertake the function of a DOminative .»
13 We have seen (p. 12) that Biurtigi (DrIeyl Ill) and possibly atirtigi .. . (Liria, XXll) correspond to Iberian place names preserved in
classical authors: Artigi, Astigi, Cantigi, Lastigi, Olontigi, Sosintigi,
and Saltigi (MLl, Prolegomena, par. 37).
Among personal names we find Biuitibas (Ullastret) (96), Bilpstibas (Mise. 26) and Tasbarigibas (ibid. 32c) as well as a number of Ascoli names: Adingibas, lliurtihas, Bilustibas, Umargibas, and Luspangib[as] . Have these names anything io do with - tigi, - tegi ? .
Basque has this suffix with the basic notion 'place where ' something exists or takes_ place. It gives rise to a number of abbreviated
forms : -gi, -ti, -di, -dui, and -doi: ardandegi, ardandui 'wine
shop', arregi 'stony place; quarry' ,jangi 'dining room' , sagasti, sagardoi 'apple orchard', masti 'vineyard', goiti 'on high; attic' , and lerdoi
'pine grove'. Lexicographers differentiate between some of these endings but they are all variants of tegi, tigi (97) .
Returning to Iberian personal names we examine Biuftibas in
conjunction with biurtigi, If the latter means 'the biur place' (where
biur is found, harvested, mined or fabricated), then Biuhibas is the
one from that place, or the owner, ruler, or simply inhabitant of that
place.
(94) R. LAFON : fEzpresslon de I'auteur de I'action en Basquea. BSL, 55, 1960, rasc. 1.
pp. 186-221.
(95) A, MARTINET : Elemen18 of generalllnguistics, London, 1964, p. 179.
(96) J . MALUQUER DE MOTES f M . OIJVA: «El nuevo plomo iberlco de Ullastret». Pyrenae, I , 1965, pp. 124 If.
(97) AZXUE says -di indicates 'abundance', -ti 'f!'equence ', and -gi 'place where'. L,
MICHELENA: .Voces V8SCaD, Em&rite, 17, 1949, p. 201 , lists all the variants of - tegi. R:
MENENDEZ PIDAL : cSobre laa vocalell ib6rlcas e f 0 en 105 nombres toponimicos., Rn, 5,
1918, p. 235 ff'.. deals with -lol, -doi as a variant of toki.
- 506 -

[page-n-507]
IBERIAN AND BASQUE

"

The Ascoli list gives us Luspanar and Luspangibas, two names
varying only slightly, possibly the 'man ofLuspan' and the 'man from
the place where Luspan is found '. The same contrast exists in Urcebas
(Misc. 53 a) and Urgidar (Ascolil. Finally a clear example of a man
from a known place is Illurtibas (Ascolil who must be from Illuro (older form, Ildurol, a settlement close to modern Matar6 (981.
We do not regard the places where something is found as being
necessarily names of towns. Furthermore, even well-known town names were in all probability based on common nouns or adjectives.
The root ildu(r), which we have just seen, occurs in so many combinations (See p. 41 that it must have a wider significance than a limited
geographical reference. U(d)urtibas is then 'the one from the ildur place' . The root of lltirda forms part of bastesiltirte (Misc. 65), iltirbigtsen (ibid. 42), and auauniltirten (ibid. 22), and it is bard to believe
that all these combinations refer to the city (99).
In fact Ulrich Schmoll, on the basis of wolves pictured on coins of
Iltfrda and Iltii-aca, decides that Utir means 'wolf as a common noun
UOOI. It is conceivable that the 'place where' can allude to an event,
as possibly BoutintibaS means 'the one from the place of victory'
(where a victory was won) if the name is based on Celtic boudi 'victory' (Holder, cols. 497-4991.
Basque toponyms ending in -tegi and -gi are abundant and the
derivatives of the same formation have been revealed in the Pyrenees
Orientaies and Catalonia (101) as well as in the old Kingdom ofValencia (1021 . This evidence confirms our interpretation of the Iberian suffix -gi, - ti as 'the place where' and provides another reason for
seeing a relationship between Iberian and Basque.

j98) M. RIBAS BERTRAN : eEl poblado iberlco de Ilduro-. Excavaciones Arqueol6gicas
en Espa1la, 30. Madrid, 1964.
' j99) PIa BBLTRAN : dos text.os ib6ricos de LW..., RVF, 3, 1953, p. 155, concludes that iJ·
dr is not a geographical term but (UIla palabra comun muy abundante..
jlool U. SCHMOLL: cOie Wortstimme Udr und ildu in der hispanlschen Namenbildungt,
Die Sprache, 6, 1960, p. 49. On Iberian ceramics the wolf is frequently depicted. It must ba·
ve had a meaning as a totem or a symbolofferocity in combat. J . M. BLAZQUEZMARTINEZ :
,Religione, prlmitivu de Hispa.nJb, I, p. 11 , mows that the wolf was associated with a god
of the underworld. As early u 1948 GOMEZ MORENO saw tha possibility that iltir and
fW'olf. were related jMlsc., p. 2781.
11011 8 . GUITER: cToponimla vasea en los Pirlneos Orientalea. APL, XIV. pp. 254-5;
d.es parlen preromans des Pyrinees Orianta1ea. Service des Archiveljdes Pyr. Or.l, Perpig·
nan, 1964; and J . COROMINAS : d.a toponymie hispanlque preromane at la servlvance du
buque juaqu'au bu moyan AgQ. Onorn8stica, I. 1960,' pp. 105-146.
(1021 M. SANCHIS GUARNER: IlDtroducci6n ala hlstorla Ungillstica da VaJend .... p. 43
fCtli" 'I1ri8'1. p. 44 fCaroigl.
- 507 -

[page-n-508]
"

W. PAmsoN

14 As we go through lists of men's names we discover that quite a
few terminate in -es and -er, and that this last dissidence seems DC·
casionally to become -aT. For example. the Ascoli bronze yields Agirnes, Arranes, Albennes, Belennes, and Enneges; another classical
source (103) gives nerdes; and documents in Iberian script record lciting only those caSBS where we can be reasonably sure that we are
dealing with a name) ildubeles (Mise. 40)' ulticeles (ibid. 8), and .. .stanes ... intanes ....banen (ibid. 15) (104).
The -er ending is clear in Nespaiser, Atanscer, and Sanibelser (all
from Aseoii), as well as Iberian inscriptions belasbaiser (Mise. 23) and
balduser (ibid. 68 and 119). These are surely men's names, and the
same may be true of sergiter (ibid. 105) and benebetaner (ibid. 67 and
70), as well 8S Utirarcer (MU, 25).
The dissidence -r is not confined to personal names, witness Sosintaker (Canet). In Basque we find the dissidence -r giving a subs·
tantive element needed to form a pronoun in cases like nor 'quien' as
contrasted with noiz ' cuando' and non 'donde' or zer 'quien, cosa'
contrasted with zein 'cual', The use of final -r to substantivize roots
is paralleled in Iberian and Basque,
That - ar is sometimes a variant of -er is not improbable, although we have difficulty in separating it from the possessive - ar, A
form like anaiosarenwi (Ens. PI. 65, nos. 11 and 13) can hardly end in
the two genitives -ar and -en, It seems more plausible to take the '
-ar as the masculine dissidence of the name 1105), Possibly d.usgitar
(Orleyl ID), balcar (Misc. 9) and carcoscar (Solaig) are men's names
with this ending. Michelena (Pirineos X (! 954) p. 443) finds that the
Aquitanian - tar is confined to masculine names , Egiar, a fairly frequent morpheme which we shall study later, is an obvious member of
this group.
Turning to Basque we food names of animals in which the ending
-ar differentiates the male from the female, Thus while asto means
burro in general, astar is the burro macho and astama, astana is the
burra, Similarly, contrasting pairs are oillar 'gallo' and oiao 'gallina' ;
mando 'mula', mandar 'mulo' ; zozar 'mirlo macho' and zozama 'mir10 hembra' ,
(1031 SU.IO !TAIlCO, XVI, 566, 571.
(1041 GOMEZ MORENO completes and corrects the first of these names to basconeS'
which has led to much speculation, especially as to ill allusion to the tribal name Vascones.
In our opinion we have a typical situation: a deceased man's name followed by his father's.
(105) TOVAR: ELH, p, 19, sees -arm.s the article plus the genitive, as in Basque Miquelarm4.
-

508 -

[page-n-509]
IBIJUA!i AND BASQUE

"

Finally, two endings indicate a woman: Bileseton (CIL, n, 3537)
and Sergeton (ibid. 2114) are both women's names (l06). Likewise
- unin is a feminine ending: Galduriaunin (CIL, n, 5922), aredaunin
IMisc. 731, cabeunin libid. 751, nisunin libid. 571, and sicounin IAlmagro, Ins. Amp. p. 72).
15 This is all we can say of Iberian morphology at present. The
omissions, such as our failure to identify signs of the negative and interrogative and even the conjunction 'and' are striking. We turn now
to questions of vocabulary, especially possible Iberian and Basque
cognates.

ill

QUESTIONS OF IBERIAN VOCABULARY

'16 If we know what a given document is about we have. much
greater chance of reaching"valid conclusions about the meaning of some (not all) of its words. The tombstone epitaphs would not contain
the same vocabulary as the tabeUae defixionum. Contracts or financial accounts will usu$lly reveal themselves by the presence of numbers. Religious topics are not easily identified, but pictures on vases
can give a hint of the meaning of the accompanying inScriptions. Pictures of dances, winged anthropomorphic figures, and animals, either
monsters or totems, imply religious significance, but unfortunately
most of these do not have accompanying inscriptions (l07). Simple
grafiti on pots and similar objects often are the names of the owners.
If keep these facts in mind, and at the same time take into consideration the morphological elements of our working hypothesis, we
may come to results which will elucidate some meanings.

(1081 MICB:BLENA: Plrineos, 10, 1954, p. 443. Cf. ELH, I, p. 383 and J . CARD BAROJA :
tComparaciones vhOO·ib6rle .... rust. Esp., dlrigida por R. MENENDEZ PIDAL, I , 3, p, 806.
(l071 The lJrla vue painting, show dances, Corpus Vasorum BlIpanorum. IJrla, LAwnu xxm (1188 Mise. 581, un, LXIII and LXVn (see Mise. 551. The flnt and last bave accompanying lnsaiptiODl. A winged. figure, Ibid., Lam. LXXD, and mODltrous animals also appear, lAm. UV, LXVm, LXXI. and LXXIV. J . MALUOUBR: HlSl Esp. dirlgida par MENENDEZ
PIDAL, I , 3, p. 322, bellev. that many of these dance scenes are cromerlu 0 prooeslonu reHglOsaD to some of the Iberian shrin... J . M. BLAZOURZ MARTINEZ : IAponeclonn al estudlo del.. reUgiones prlmIdvu de Eapa4P. A. E. Arq., 30, 1957, p. a., conftrma this bellef.

-

509 -

[page-n-510]
"

W. PA'I"I'ISON

17 Many scholars have. suggested that certain Basque words may
have Iberian origins (108). A few ill-advised attempts have been made to interpret whole Iberian dotuments by Basque (lOg) but the accepted opinion is that the etymological method, starting with Basque.
will give meager results in Iberian. The combinatory method has been
restrictcj up to recent times because of the paucity of documentary
evidence. It is this second method we hope to employ to the limited
extend of !·s possibilities, with occasional glances at possible Basque
countt. ••' dI'ts.
18 The first tablet found at La Serreta (Alcay) has the words amai
sakarisker written transversely across one end of the principal text,
something like a final thought or 8 signature added to a letter which
has already fIlled the page. In all probability sakarisker is a proper
name or title; it reappears in the 'lead' of IJria (Mise. 74g) and seemingly in the Tartessian inscriptions sacal iscer (Castulo, modern
Cazlona) (110) and sacarbiScar (Mogenta) U 11),
The fIrSt part of the name recalls the root sacr- found in Latin sacer, sacerdos and in Celtic names (112) such as Sacerus, Sacra, and
Socer,
After the Roman occupation of Cartago Nova one Qf the magistrates of the city gives his name as Sacerdos (113) which looks suspiciously like a latinization 01 sakarisker. In any case, the root sacrmay be related to 'sacred' and 'sacrifice', or as G6mez Moreno (Mise.
p. 281) says, «todo inelina a sospeehar que 81 saeariscer sea nombre
de entidad sagrada,)

(108) A partial list : A, TOVAR: lE! au.skera Y SUI parlentep, pp, 38-56; ELB, I , pp_ 18·21 ;
P_ BEtTRAN: l'Textos ib6riC09 de IJri8.l. RVF, rn, 1953, p. 41 0'.; D_ FLETCBER: Die Spracbe, 16, 1970, pp. 167-8 ; L. MICHELBNA: Em6rita, 23,1956, pp. 277·280; ArchivUDl, vm,
p. 43 and Plrineos, X, 1954, pp. 410·443 ; GOMEZ MORENO : Mise., pp. 278-281: R. MENENDEZ PIDAL: l'Toponimla mediterdnea y toponimla valenclana prlmitiVh. Bol. de Dlaleetologla Espaftola, 33, 1954/5, pp. 61-75, and J . CARO BAROJA : op. cit., note 106, pp. 789 fr.
(lOg) Por example, L. GALERA: cBlltlayo de lectura del plomo de El Solalp In APt. 13 .
1972, pp. 127· 137 .
(1101 A. VIVES: n, pp. 168 and 172.
(1111 G. BAHR : op. at., note 62, pp. 425 and 427; PIa BRLTRAN : up. at., note 38, who
transcribes (p. 36) this word a8 saJuubisgabe. See below p. 41.
(112) A. HOLDER : n , cols. 1275 and 1279 ; J . WHATMOUGH: d>lelectsof AndantGaw.
Cambridge, Mass., 1970, pp. 338, 343, 352. and 653.
(1131 A. VIVES: up. cif., IV, pp. 33 and 36 and avlli oftbe prologue ; A. BRLTRAN : .sobra la moneda de Carthago Nova con ·SacerdOS't. A. E. Arq .• 20, 1947, pp. 137.141, inclines
to the opinion that Sacerdos is not a name, u Vives believes. For another example of Sacerdos as e name, see J . VIVES, DO. 1357.
- 510 -

[page-n-511]
IBERIAN AND BASQUE

25

Strabo {114l calls the Lusitanian priest a 'Hieroskopos', who divines the future by examining the viscera of animals and sacrificed prisoners. Similar rites occurred among the tribes of northern Spain
(115l. Ancient cults continued their practice in the west and north,
but those of the east disappeared under the influence of Greek and
Roman religion at the beginning of the historical period. It is not surprissing, since almost no name of a primitive god is found in eastern
Spain, that records of pre-Greek and Roman religious practices
among the Iberians are rare. But in the earliest Christian times. lingering pagan superstitions were condemned, including the divining of
the haruspices who sought their omens in the entrails of their victims (116).
The root sacar leads us to the Basque sakarramin 'entrails' ; sakar
'rubble; pus' , sakaila 'big wound'. It seems that a root (sacr-) having
to do with divining the will of the gods through entrails has been
adopted by both Iberian and Basque (or passed on to Basque through
Iberian?).
19 The second part of sakarisker sometimes takes the form esker
(standing alone , Misc. 37c and 119), in Aiunesker (Mise. 32a) and
... urcescer (La Serreta IVI. It seems to form part of the Latinized names Tanneg/iscerr/is (CIL, II, 3794) and Baes/isceris (ibid. 3221) ; also
in shortened form, in the name Atanlsker in the Ascoli document.
Where we also find Arbliscar. Iscer can also appear as the first element of a proper name: Iscerbeles (Misc, 100) and the Tartessian Isceraden (Mise, p, 267).
Apparently isker, esker has not usually been related to -esken,
found only on coins, Traditionally taken as a genitive plural with relations to Basque, this latter sufflx has more recently been called an
ethnika (117) simply indicating that the coin in question is from such
and such a city, Among the many examples Tovar cites: arsescen, ausescen, iltircescen laiescen, otobeScen, and undicescen, as well as the
Tartessian icaloscen and urcescen (118), It seems clear that if -esken
11141 A. GARCIA Y BEUJ.I)O: cE9pafaa y 108 espaAoles hace dos mil &A05l. Madrid, 1968,
cuarta ed" p, 118 ; J , M . BLAZOUEZ MARTINBZ : .Religiones primitivas de ESp&Ab , C.S.I.C.,
1962, p. 23,
(1151 BLA20UEZ MARTINEZ: op . cit" p . 32.
(I 16) M. TORRES in Hist. Esp., II, pp. 452 and 479, n. 49,
11171 A. TOVAR: Lex., p. 317 andELH, I, pp. 24 and 18, n. 34, where he speculates that
- escer may be an ethnica related to -escen; he anali%es -escen as a compound suffix, like
the Basque instrumental -z plus the Ind06uropean - ko.
(I181 On - sken, see SCHUCHARDT: Iher Ded., p. 31; J . VALWO: ISobre iberico - sken
y -en ), Emerita, 18, 1950, pp, 215-220; MICBBLENA : .Sobre el estado actual de la cuesti6n del genitivo vasco en-eru, Bm6rlta, 18, 1950, pp. 221 -224; SCRMOLL, op. cif .. note 67.
p. 62 ; A, TOVAR: Lexlca, p. 291 , under arsescen, and p. 317 under -scen.
-

511 -

[page-n-512]
26

W. PAT1'ISOlf

is an ethnic this suffix cannot be related to isker, esker, for Iscerbeles
defies interpretation as a localization.
Taking isker. esker and -esken as related to the Basque esku we
come to a plausible solution not entirely at odds with Tovor's idea. Es ku
is fundamentally 'hand', but among its meanings we fmd 'derecho,
fsculted' which idea dominates in eskuantza and eskubide (both mea·
ning 'facultad, derecho, 8utorizaci6n'), 8S well 8S in eskudun 'minis·
tro, persona dotada de autoridad' and eskuera 'juris4!pci6o', (119).
A coin, then. inscribed Undikesken is issued 'by the authority of
Undike ' (or Indike) and Sakarisker is 8 man 'empowered' to 'sacrifice'
(or with authority in religious matters), in short, a priest. Escer seems
to be the equivalent of 'magistrate' or 'official', The combination ur·
cescer (La Serreta IV) is '(the) magistrate of Urei'; aiunescer (Mise.
328) '(the) aiun official', where aiun (120) is not necessarily a place
name, any more than sakar was in sakarisker. These combinations
with esker could with time become simply personal names, as Priest,
Bishop, Pope, and Mayor have done in English, as well 8S Alcalde,
Conde, Reyes, and others in Spanish. Thus the magistrate of Cartago
Nova named Sacerdos may have been the son or grandson of a priest,
rather than a priest himself.
An inscription from Liria around the rim of a pot cover reads
...giscer egiar bancebereiwbar baIduser ban (Fletcher, I . 1. LXXVI)
which has something to do with a noble (egiar) official (iscer) , son of
Balduser. Egiar, I hope to show, is something like the Spanish 'senor' ,
and is appended to many names in Liris inscriptions.
The last example we shall examine is lskerbeles on a coin also
bearing the inscription Undikesken (Misc. 100). Hence lskerbeles is
indisputably a magistrate's name. Beles, bels, has constantly been related to Basque belu ' black', but in Iberian, where it occurs frequently
in men's names (121), it may have had a more eifU'ID.stive, positive
value: perhaps something to do with the moon goddess whom the Romans called Hekate to whom they sacrificed ~lack puppies and black
lambs and who presided over magicians and enchanters. The Iberians worshipped the moon goddess on the island off Mainake as well

t119) LaUD manus also hu 'power' at one meaniDa.
Two coinl bear !eaands endina in -cm inltead of __,cm, which may be an abbreviation,
a variant, or a acrlbeJ error.
1120) 'Aiunl' appean OD a tombstoDe (Mile. 441.
11211 See TOVAll: L6x.. p. 296 tunder bele,).

- 612 -

[page-n-513]
IBEIUAN AND BASQUE

,

"

as other points along the M.editerranean coast (122). She is perhaps
the nameless Basque divinity, fearful yet beneficent, for whom the
Basques danced on nights of full moon (Strabo, Ill, 4, 16). Iskerbeles
would be '(the) man empowered with magic' or simply 'the magician'
(123). His name recalls the Basque belhagUe, belhaile 'brujo, hechicero' which seems to combine bel 'negro', ageri 'declarar, nianifestar',
and -gUe, -Ue the ending denoting the actor or agent; hence belhagile
would be 'the one who reveals the black or occult'.
We must record, however, the fact that Michelena (124) derives
belhagile from belhar 'hierba', giving its meaning as 'bruja', lit. 'hacedor de hierbas'.
As for the final letters of -sken and isker, esker, the dissidence
-en appears to be the same which Basque uses as a possessive, and
which was widely diffused in place names throughout the Iberian region. See above, p. 10-11. The ending -er shows up often, in addition
to isker, esker. For the -er ending, see p. 22.
20 We are fortunate to have a learned work on the geographical
distribution of personal names in ancient Hispania (125). As might have been supposed, the elements isker and beles are confined to the
Iberian region. When, after the Roman conquest, Iberians adopted
Roman names, it was only natural that some of them translated their
Iberian names into Latin. We find Pollio 'Powerful' and Niger 'Black'
appear frequently in the Mediterranean coastal area and in Andalucia
and Portugal. Only 8 few occurrences of these two names are recorded for central and northwest Spain. Does Pallia translate isker as Niger does beles? Furthermore, can we relate Porcius, common in the
Iberian district, to the tribal name at the Cerritani, famous for their
hams (126), whose name is thought to relate to tbe Basque zerri 'hog',
cherri 'pig'? We also wonder if other Latin names peculiar to the Eas-

(122) A. SCHULTEN : Geogr., I, 323, and n. 43. In Celtic territory a god represented by a
crow and whose name contained the syllable lug (Indo-European 'black') is catalogued by J .
M. BLAZQUEZ MARTINEZ: !Religiones primitivas de Hispanill), I, p, 89 ,
(123) Another poaalbUity is that beta is related to Celtic Belenos (the equivalent of Apollo), derived from the Indoeuropean ghel 'Wanco, brillante'. In Iberian territory we find Belenos in Azalla (M.lsc, 32d) and Belennes in the AscoU Ust,
.
(124) MICHELENA : cHisp6.nfco antiguo y V8lC0. Archivum, 8 , 1955, p. 40, n. 15.
(125) J, UNTERMANN : dUementol de un atlas antropon1mjco de la Hispania antigull),
Madrid, 1965,
.
(126) See SCHULTEN: Estrab6n (in Fontes Hispaniae Antiquae, VI, p, 246), and Geogr"
ll, p, 513.

-

"

513 -

[page-n-514]
20

W. PATTlSON

tern zone hide native Iberian names of which they may be transla·
tions, for example Granius. Grattius, and POStumiU5 (127).
21 We have mentioned Urcebas (Mise. 53s). Urcescer, and Urgidar (Ascoli), three men whose names derive from Urci, Urgi, recorded
as a place name by Mela (IT, 94) and Pliny (N. H., rn, 10) {128). Tbe
latter also speaks of Iliturgi and Isturgi. Furthermore, Latin inscriptions bring us Urcestar (elL, 1I, Dr. 2067) and Urcico (ibid. Dr. 2818).
Urke appears without suffix (Alcoy, La Serreta 11 and in the compounds urcetices (Mise. 261. urcecerere (ibid. 43) and Arcedurgi (MU,
23). Apparently more than a couple of places called Urci existed1n ancient Iberia.
Caro Baroja surmised that Urci means 'fortress ' , basing his
thought on Pliny'g words cUrgia cognominata Castrum Julium» (CB,
808). Consequently Urcebas could be 'the ODe from the fortress'. But
we must not overlook the·Basque urcia 'God' according to Aymeric Picaud (129), and the modern Basque ortze 'sky, Heaven', ortzadar
'rainbcw' {literally 'sky arcb'). and ortzegun 'Tbursday' , Ille day of
Ille sky god, just as Jueves is also '(Ille day) of Ille sky god' (130). A
usual modern Basque word for God is Jaungoikoa 'the lord of the high
(place)' or 'of heaven'.
We think that Urcebas must mean the 'one from the height' and
Urcescer can be the 'magistrate of the high (town)'. It is very possible
that a number of settlements - as we know, generally built on high,
easily defended places - should have 'high' or 'height' as their names.
Hence, we suggest that the fundamental meaning 'height' be ascribed, at least tentatively, to urci (131). A secondary meaning 'fort' or
'acropolis' is not precluded. There seems to be a correspondence between urci and Latin arx, arcis 'fortress, castle; a height; a mountain
peak; (anything) high'. The two seem to be combined in the unidentified city's name Arcedurgi.

(127) Granlut and Porcius gava rise to the place names Granena and Purchene, MENENDEZ PIDAL: op. efl., Dota -48, pp. 20 and 27.
(l28) See above, p. 18; J . UNTXRMANN: dtstudio sobre Iu ireu Iingillsticu prerroma·
nas ...•. APL, 10. 1963. pp. 187-188 and map 9.
(l29) Tbe 12th century author of • guidebook for pllgrimt: to Compostela.
(130) Several mountains called Mon, Jom (modern Mong6) bear wltn8S:B to the wonhlp
of the eky god 10 the high places. SCHULTEN : Geogr., I. pp. 325, 328 and 330.
(1311 SCHULTEN : ibid., I, pp. 219-221 ; and Hilt. Bap., I. 3, p. 324, describe Iberian
towns and shrines u built on high places.
-

614 -

[page-n-515]
IBERIAN AND lIASQUE

29

22 Besides escer, iscer two other recurring words seem to refer to
the position of esteem enjoyed by the recipient. Egiar, which may
stand alone or be affixed to the preceding word, is particularly com~
mon on the painted Liria vases, where human figures, often warriors,
are depicted. In some cases there can be little question that the words
refer to the man, as for example caresbobigir egiar inscribed between
the horse's legs under a mounted warrior (Mise. 53b. with reproduc ~
tion of the vase; Liria, plate LIT). On the same vase fragments we find
carestirte egiar and ebirteegiar which have. however, no clear reference to specific men in their placement. Another fragmentary pot
with pictures of horsemen, women, and musicians (a ritual or a triumphal parade?) bears the words .. .rbancusegiar biurtitee[qiar1 besumincuegiar (Mise. 55, Litia, plate LXVII, Dr. 7). Still another frag~
mentary vessel, painted with war and hunting scenes, has next to two
combatants cemiegiar and ecuegiar (Misc. 56; Liria, plates 48 and
49; reproduced in M. Pidal, Historia de Espana, I, 3, p. 624). Finally,
a coin of Arse (Sagunto) is inscribed Arsagiscuegiar (See above p. 11)
which must be the title of the official who issued the coin.
Several scholars have seen a relationship between egiar and the
Basque verb egin 'to make, to do' (1321. This verb is regularly used
with nouns to express physical actions (as opposed to emotional states). For example, 'to sleep' is 10 egin, 'to cough', eztul egin. lt enters a
compound indicatin.g one who deals with or works with something :
legegin, 'legislator', harrigin, 'stone worker'.
We take egiar as a 'doer' of deeds of importance. whether it be in
the military or governmental sphere. This interpretation is supported
by the appearance of the word with warriors and on the coin of Arse.
Hence we equate it with 'chief, leader, lord' and 'official' . It may ha ~
ve about the same semantic content as the Latin eques.
For the ending -ar on the end of egiar, see above p. -22.
A second word which may be a title of some sort is iunstir. ltis re ·
corded fairly often, sometimes as iumstir and also as iustir. Signifi ~
cantly, in the 'lead' of Solaig, it stands as the fIrst word of the two lines of the inscription: iunstir belesair and iunstir egiartone. In the
'lead' (Serreta I) of Alcoy iunstir (written iumstir) is the first word of
the second text, comparable to iMke (which may be the Celtic riks, riki
'king' ) that begins the first text. Although we advance our suggestion

(1321 TOVAR: Ux., p. 306, mentions BELTRAN, CARO BAROJA, GOMEZ MORENO.
who support this belief, and adds his own approval.

-

515 ":'

[page-n-516]
30

W. PA'M'lSOH

with reserve, it seems probable that iunstir is a title, especially as it
begins what may be proclamations issued by noblemen. The fact that
it is accompanied by egiar in some of the texts strengthens our notion
that it is a title (133).
Cuadrado saw a relationship between iunstir and the Basque iaun
'lord' (134) this root appears in the name cL. lunius Iaurbelesl from
Guisona IUrida) (J. Vivas, no . 2492; HAE, no. 496), which can be

compared to Basque derivatives of jaun such as jaurgo 'sefiQr1o' and
jaunri 'gobernar, defender, amparar'. Iaurbeles shows us that the
root iaun was known in Iberian territory, thus adding to the credibi·
lity of a relationship between iunstir and the Basque root. Since the
latter root meaDS 'senor' our belief that iunstir is an honorific title is

supported.
23 We now turn to other vocabulary item which may be related to
Basque cognates. It goes without saying that much of what follows is
only tentative.
Al An inscription around the rim of a Liria vessel (Mise. '54; Fletcher 1. I., Insc. IXl reads: bancurs caresbanite egiar saldutibaite
iumstir tolirlbitane bas.rumiwtinire. The painted figures on the vase
represent two horsemen, galloping in the same direction (See Liria,
CVH, p. 37 and LAm. xxxivl. Aside from the two riders with their horses no other man or animal is depicted, which causes us to conclude
that' the inscription must refer to these two persons.
Now both egiar and iumstir can be taken as titles of rank or nobility, something like 'lord' and 'prince'. Caresbanite, which we divide
cares-ban-ite must be '(thel cares ones' and with egiar 'the lord of the
cares people' (135). Similarly, saldutibaite iumstir becomes 'the prince of the ones from saldu'.
B) The Iberian -aur has been equated to Basque aur 'child' (see
Lex. p. 2881, expressing filiation in the compounds atin/bel/aur (MLI,
vi), laeer/bel!aur (Mise. 16) (1361, oreeieelaur (BinMar) and belagasi·

(1331 FLETCHER: Die Sprache, 16, 1970, p. 153 , comments OD the combination ofiuns·
tU- with egiar and also bitan.
11~41 See FLETqrBR : ,vUlares VI, nuevo plomo ib6rico escrltoJ. APL, pp. 196· 7. for bibliography and interpretations ofiunstir. In medieva1 Buquejaun la a frequent proper name.
MICHELENA, op. cit., DOte 89, p. 49.
(135) Compare clU"mnll egW- (Mise. 5381 on another Liria pot.
(1361 Compare lacereiarduru (kisc. ll81 and l.acmlis (pen., CIL, n , 46251.
-

616 -

[page-n-517]
lBlIJUAR AND BASQUE

31

kaur (Serreta I) (137). If these words can mean 'son of Atinbel', etc.,
bow are we to interpret aurbiu/r) and aui-Iso ... (both in Ullastret, plo·
mo 2)1 According to our bypothesis aur- is the determinant, hence it
bas an adjectival function and would be 'youthful' or 'childlike', if de·
rived from Basque aur. Further doubt is cast on this interpretation of
-aur by the study oftbe inscriptions of Pico de 105 }\jos byD. Fletcher.
He finds the division bel·aur is unlikely, as laur, without be-, is more
frequent.
C) Among the radicals we isolated earlier was baite(s), baiti (see p.
6). Except in the form baitesbaniecarse (138), where it occupies the
initial position, baite, baiti acts like an adjective, following another
radical or standing alone and taking derivational sutnxes. Basque
baita (Azkue, 3rd meaning) is described as a particle added to a name
to designate the house of the owner, but most commonly as an inflX
between a personal name and a suffix such as -n, -rik, or -ra. Bas·
que baiten 'in, among' shows this root combined with the inessive en·
dingo The occurrence of the same word in northern Italy leads to its
classification as a remnat from a substrat language (139).
If we take baite(s) to be ans adjective, and assuming that its mea·
ning is related to Basque 'house' , we would have to interpret it by 'do·
mestic; home·like' or perhaps 'familiar'.
D) Sosin is often a component of men's names; Sosinasae (Ascoli).
Sosimilus (ibid. < Sosinbiiis). Sosinaden (ibid.) and Sosintaker (Canet).
But other uses of sosin lead to the conclusion that it had a meaning as
a common noun or adjective. Sosintigi (Pliny, N. H., m, 15) a place
name, and (s] osintiba (Fletcher, Bronce) seem to refer to a place
where something called sasin was found (although perhaps it is the
place of residence of a man named Sosinl. Sosinbiuru (Misc. 43)
shows sosin applied as an epithet, although it could still be the proper
name as a modifier, just as in Sasintaker, 'Sosin's place or grave', In

(1371 La.. certain examples are argiboti-b«aur (LAFON : op. at., Dote 501. dur$/aur/ar
(Mise. 71; ALMAGRO : Ins. Amp., p. 76, suggests dJ.us/biur/atu and balcebaur (so giveD by GOMXZ MORENO ; Mise., p . 324. whJch we read as balcesur from JANNORAY: Ens., Plate
LXVI, 281. Also see D. PLETCHER: d.os plomos iberiCOIJ de YAtova (Valenciat.. T.V.S.I .P., 66,
Valencia, 1980. pp. 17·8.
11381 See below p . 38, for speculatioD OD the possibility that baites· banl«arse is a verb.
11391 TOVAR : cTbe Buque LanguagM (l:ransIated by H. P . HOUGHTONI, p. 72, points
out that baita. baitha means 'house' in Guipo.zcoa, in dialecu of North Italy, Langue d ' Oe and
GascOD.
-

517 -

[page-n-518]
"

W. PA'M'tSON

nabarsosin (Almagro, Ins. Amp., pp. 72 and 260) we are tempted to
see a relationship with Basque nabarben, nabargarri, nabannen 'ex·
traordinary, outstanding'. Since according to our hypothesis Ip. 101 nabar in the initial position of the compound must be a noun, we have
to postulate a meaning something like 'prince, nobleman' .
Micheleoa has suggested a relationship between sosin and Basque
(h)osin, 'pozo.lugar de mayor profundidad del rio' (140) . Ifwe accept
this meaning we can assume that the mao's name was chosen to inspire awe and fear (see above p. 61. Names of an awe-inspiring nature are recorded in many other cultures, and are natural in a society
given to warfare, like the Iberian.
El The r:oot tigir, ticer. tigis, tices may represent the Celtic tigernius, tigernus, tigirnos 'HeIT, Kenig' (Holder, cols. 1841-21. Thus a
name like isbetarticer (Mise. 471 would be 'Lord Isbetar'; alortigis
(Mise. 39) 'Sir Alor', and urcetices (Mise. 26)' sovereign, supreme
lord', Argitieer (Mise, 43) 'lord of light'1(141J eould stand for a god,
for instance a sun god like Apollo.
Another possibility is that this root is present in Ticer, Ticis, ihe
name of the river near Ampurias (modern Ter) recorded by :PUny, N.
H., 3, 22 and Mela 2, 6, 89 (1421 . The well-know worship of rivers
could lead to the name 'sovereign' given to this one.
24" We have said nothing about possible verb forms and what we
have to say reveals little that reminds us of Basque. It is true that
-dedin in bidudedin (Serreta I) and sesdirgadedin libid.) and dadula
(ibid.) do have the appearance of Basque forms. But the other words
which we tentatively classify as verbs show no resemblance to Basque, except that they seem to be able to compound a noun and a verb
just as Basque does in combinations like aurduna 'the (woman) that
has a child' and arrigina 'the (man) who works with stone'. It looks as
if two words in the 'lead' ofCaste1l6n (Misc. 43) are such compounds:
astelbeiceaie 1and aurunilbeiceai. They can be compared to arnai
{Serreta 11 turlbai {ibid.}, iSbinai (ibid.) and ilduniraenai (ibid.) which
may be verbs. Castell6n gives us another compound with a possible
verbal ending: balcebiuraies (Mise. 431. The same kind of compounding may occur ingoloitecari (Mise. 761 and berbeinari (ibid.) both of

(14-01 MICHELBNA: tCulHltiones relacionadu con la escritura ib6riCb. Em6rita, 23,
1955, p. 279.
1l4IJ BAHR : op. dt., note 52, p . 419.
(1421 See HOLDER, col. 184-0.
- 618 -

[page-n-519]
IBERIAN AND BASQUE

33

which are followed by eugiar, possibly related to Basque euki 'tener;
poder {substantivol' just as the more common egiar has been related
to Basque egin.
Other possible verbs end in -se: ultiteceraicase (Misc. 43), bencarsense (ibid.) and baitesbaniecarse (ibid.). It is noteworthy that all
these supposed verbs appear in the long inscriptions, whereas in the
brief epitaphs there is no need for anything but nouns. The long inscription of Cigarralejo (Mula) lacks word dividers and it is difficult if
not impossible to isolate possible verb forms in it.
25 Numerals and measures. We have called attention to the nota tion of numbers (up to nine in the. lead o.f Glidor) by means of venical
strokes (p. 151. In another case lLiria, lead % I1 we find 22 such strokes, but this we take to be a running accoint where one stroke was
added each time another unit of whatever was being counted was added. The writer did not know in advance how many units were going
to be delivered.
It is different in cases where the total number of units was at
hand from the beginning. Now the scribe could sum part of the digits
into larger numbers. as we might combine them into tens and hundreds.
This is the case in several inscriptions where the total is divided
into A. O. and Ki. each followed by a certain number of vertical strokes. Examples: Sagalaguga AI 01 Ki I (Serreta VI. cara A). and unwsa
aren A 1 04 Xi 4 (La Gra~uela) (143); in addition there are other inscriptions using only part of these symbols. such as A 1 0 11 I 11 {Santisteban del Puertol and 0 111111 Ki 11 (Serreta VI, cara B111441.
Aside from Tovar's identification of A with sextarlus. H with hemina, and Ki with kyatbi, to which there are two objections (145). no

1143) Publiehed by TOVAR: dnBc:ripci6n iberlca en una 8amella del teeoro de La Granjuelh. RABM, 61 , 1955, pp. 580-583 and ELH, I, 19. TOVAR takes the symbol we transcribe
by 0 to be H, whlcb le colTtlCted by FLETCHER: .orleyl Illi, A. E. Arq., 40, 1967, p. 56. This
epoUs TOVAR's interpretation of H as hemina.
(144) Serrata VI continues below the cited numerals with 0 11111111 Id Ilri 11 A 111111, another
case, we think, of a running account wbere new figures were addad at later times. FLETCRER : op. dt .• note 143, p. 55 ; Die Spracbe, 16, 1970, p. 159 ; andAPL, 13, 1972, pp, 120-1 ,
deals comprehensively with numerical inscriptions of the type we are considering. FLETCHBR's recently publiabed d.oe plomos ih6ricos de YatoVh (see note 137) contains another
type of numerals of great interest.
(1451 First, the confusion of H and 0 ; second, the fact that If the second element is four
beminu It wowd be more than ona sextarlus, u observed by FLETCHER: A. E. Arq., 40,
1967 , p. 55,
- 519 -

[page-n-520]
.

W, PATrlSON

to my knowledge has tried to give values to these symbols. We
suggest that we have here the Roman system in which A is as, originally 8 pound, 0 equals oncia, uncia, 1/12 of the as, and..ris the sign
of the sextula, 1/6 of the uncia, or 1/ 12 of the as. The last symbol
brings in doubt, as the dimidia sextula, 1/12 of the uncia had a sign t.
which resembles even more closely the simbolSs we trsscribe Id (146).
According to this system the total contents of the vase of La Granjuela equals ODe as plus .333 as (4 unciae) plus .0556 as (4 sextulae),
giving 1.3886 as. Since Tovar tells us that the vase holds 1.715 liters
when filled to the ring on the neck, the unit A (or as) would then be
equivalent to 1.235 liters. When the vase is filled to the top the. CODtents are 1.960 liters and on this basis the as would be 1.41 liters.
There is not, however, any ancient unit of measure which corresponds to either of these figures . Perhaps the annotation on the vase
.
has nothing to do with its contents.
In other Iberian cities different systems of measurement were
employed. A document ofl!nserune (Plate 65, no. 211 bears: e 11 1. At
Azaila we find ~TTlI A 111 (CVH, fig. 17, 30) as well as severalapparent numbers involving the S sign: 0 S S (ibid., fig. 18, 301, SS (fig.
19, Ill), SSS (fig. 19, 117), and I'\S (fig. 19, 125). Tbeserecordings
(Serreta I), whlcb Scbulten interprets as 3 stabring to mind ~h 'X
ters and 1-1/ 2 chalkoi (1471. Possibly the MM ( SS ,una XIII) and
the SS (ibid., LXmI belong with the numbers just cited. Tbe recently
published Villares V (Fletcher, cCinco Inscripciones), p. 2011 is a do·
cument of great interest for numerals as is his even more recent «Plo·
mos ib~ricos de ylttoV8» (See n. 137). They are apparently of commer·
cia! nature.
Far more engrossing is the question of numbers on the 'lead' of
Mogenta (148). This tablet, which was found under a hand mill, whe·
ODB

<

(146) See FLBTCHER: APL. 13. 1972 . p. 120 and F. HULTSCH: Grlechische und romis·
che Metrologie, Berlin. 1882 . p. 148.
The system of the as. uncia, etc., was used for the meuW'8 of weisht, coinage and liquids, although the Iigniflcance of the units varied in each cue. Also with the passage of ume the value of the units changed greatly. See M. LA.ZZAIUNI : &Metrologia roman .... Conimhriga. 4 . 1965, pp. 81 -95. I have DotbeeD able to consult the work ofF. J . DE OROZARIZCU·
UN : cElliltema metrol6gico de la intcrI.pcl6n de La Gr&DJuel.... n Coloquio sobre lenguas y
cultura prerromanas de la Peninsula Ib6rica lTilbingen. 17-19 junio. 19761. Salamanca,
1979. to which D. FLETClIER calls my auention in a letter of Mey 2, 1980.
(l47) A. SCHULTEN : cPonchUIllen In SpanleDl: Arch. A.n.I:elgar. 1933. col. 5 Z1 .
(1481 Published byPIO BELTRAN : op. elt., note 38 ; and D. FLETCHER: op. cit., note 32.
pp. 46-48.

- 520 -

[page-n-521]
IBERIAN AND BASQUE

35

re it was apparently hidden not loog before the settlement was destroyed, contains on one side a series of words separated by dots, ranging in number from one to six and crossed out by a vertical line
drawn through the dots, except in the added text, written upside
down at the bottom. (This description refers to Cara A in Fletcher,
which Beltran calls Cara B.l The text of the other side (FletcherB, Beltran A) has all the words crossed out by a line drawn through them
with the exception of two, but the dots, which number from one to
ten, have not been touched.
The conclusion that we are dealing with an account of some kind,
perhaps a commercial record, is hard to avoid (149). The crossed out
dots and, on the other side, the crossed out words, seem to indicate
closed accounts.
The problem is complicated by the fact the script of the documents is the southern Iberian, or Tartessian script, and the value of
various signs is in dispute. Particularly important is the symbol 9 ,
which in eastern Iberian equals r, but which Pio Beltran takes as be
at Mogente. This means that a word transcribed by G6mez Moreno's system as saldulacogiar (150) and which looks very much like a name
with the honorific -(e)giar added becomes saldulagogiabe (Beltrlin,
op. cit., p. 36, words 13 and 20), and nine other word.s which Fletcher
transcribes with the ending -gar become -gabe for Beltr8n_ This
leads to an equation with Basque gabe 'without' and the belief that
the whole text is a tabella defixionis, especially since one word aiduar/begi/abe is related to Basque aidur 'maligno', begt 'ojo' and gabe
(Beltrlin, op. cit.. p. 35 If.) (151).
Against Beltran's opinion we can adduce some objections_ In the
first place not even the eminent G6mez Moreno could make a definitive transcription, In his article published in 1961 (152) he revises his
opinions and gives the values 0 and e to the symbol? ' Saldulacogiar

11491 A. SCHULTEN : op, ail., col. 522 ; J . DB C. BERRA RAFOLS: cNoves inscrlpcions iberiqueSl. Institut d' EstudJs Catalans, Anuari, 8, 1934, p. 339, both support this idea .
11501 FLETCHER's cara B, line I and liDe 3; compare this word to sacalacucaegiar followed by a number lSerreta VI) and w ords formed on the radical saldu-: salduie lMU, 351, saldugi/erku lUUastret, plomo 31, and saldu1wbalkuleboberkur ... libid,l.
1151 1 AZKUE gives other meanings for aidur. If the adjective 'maUgno' is chosen, it
should not stand in the initial position. See our hypothesis, p. 6.
11.521 M, GOMEZ MORENO : u.a. escrltura bbtulo-turdetanu. RABM, 69, 2 , i 961 . pp.
879-949. This article postdated BELTRAN's work by seven years,

..

- 521 -

[page-n-522]
36

W. PA'ITISON

becomes saldulacokiae (p. 932) and rdarbauba becomes otarbeobe fp.
930). In the second place Beltrlm gives no attention to the numbers indicated by the dots or to the words stricken out.
Contrary to Beltdn we placA S'.. e8t importance on the numerical
notations of this document and feel that at least some of the words in
-ar must be personal names. Hence the document is for us simply a
record of a commercial transaction.

IV

CONCLUSIONS
As a result of the comparisons we have made between Basque

and Iberian we think: there is a clear structural relationship between
the two languages. We feel that they compound radicals with the S8me distinction between nouns and adjectives, and that this distinction
permits us to narrow the possibilities of ascribing meaning to some
radicals.
Among Iberian inflectional suffixes we believe there is sufficient
evidence to accept the meaning 'one' for - bat -ban, - has and see a
correspondence in Basque. The significance of -ti, -gi as 'place whe ~
re' and its identity with take, tegi is plausible, as is also - ite as a plu~
ral sign, including the interpretation of - tar as an Iberian plural pos~
sessive. The particles -en, -ar and -cu, - co seem to correspond to
Basque counterparts.
Among vocaoulary items we maintain that the rendering of iuns·
tir and egiar as honorific titles is a strong probability. The relation ~
ship of -escen and escer with each other and with Basque esku. seems
justified. The meanings ascribed to urci 'height' and Sacariscer
'priest' we think are warranted.
Far be it from us to claim any special type of relationship between
Basque and Iberian. We are content to say simply that they appear to
share certain morphological and syntatic structural elements. Whet·
her these similarities depend on a common origin, on a borrowing of
one language from the other, or on the influence of a substratum is im~
possible to decide. Until much more investigation on a purely lioguis·
tic basis is possible the reasons for the correspondences between the
two languages remain a closed book.

-

622 -

[page-n-523]