The Basque names for the days of the week are as follows; the Bizkaian dialect is noticeably divergent:

The major account of these names is Michelena (1971b), upon which my account is largely based, but see also Caro Baroja (1973: ch. 6).

The first three names are compounds of <aste> `week' with <lehen> `first', <arte> `interval, between' and <azken> `last'. These transparent if unexpected formations have induced many commentators to suggest that the Basques anciently had a week of only three days, an arrangement which would appear to be highly inconvenient and which does not appear to be paralleled elsewhere. In connection with this, I find it significant that these forms are attested everywhere with unreduced <aste->; nowhere do we find *<asta-> or *<ast->, even before a vowel, which suggests to me that these names are of no great antiquity. B <ilen> and <martitzen> are calques on Romance: <il-> is `month' (earlier `moon'; see the next section), and the whole is a reduction of *<il-egun> `moon-day', while <martitz> is `Mars' (apparently from the Latin genitive <Martis>), and the final element is possibly the genitive <-en> but more likely a reduction of <-egun>: hence *<il-egun> `moon-day' and *<martitz-egun> `Mars's-day'. Bizkaian <eguasten> is a compound of <egun> with an unidentified second element; a plausible surmise would be *<egun-aste-egun> `day-week-day' (see <eguen> below).

The name of Thursday has attracted an enormous amount of comment and speculation. Eastern <ortzegun> is conservative; western <ostegun> shows the familiar development of <rtz> --> <st>. What is striking is that all varieties of Basque show a number of words involving an element <or(t)zi> ~ <ortze> ~ <osti>, all of which have meanings pertaining to `sky', `storm' or `thunder'. Table 5.1 is a collection taken from Azkue's dictionary; items for which no provenance is given are widespread.

To these I can add the exclamation <Ortzi zuria!> (<zuri> `white'), roughly `Bloody hell!', which I have heard in Lapurdi but never seen recorded. Now the first element of <ortzegun> appears transparently to be this widespread item <ortzi> `sky, storm, thunder', and what has fascinated the observers is that this looks remarkably like a calque of the Latin name of Thursday, <Iovis dies> `day of Jupiter'. Jupiter, of course, was a god of thunder, and it is well known that the Germanic peoples calqued the Roman name of Thursday with the name of their own thunder god: hence `Thor's day', or `Thursday', in English, and `Thunder's day', or <Donnerstag>, in German. And from this follows the inference, drawn countless times and repeated as gospel in a number of popular books, that <Ortzi> must have been the name of the ancient Basque god of thunder or of the sky in general.

This inference gains support from the report of the twelfth-century French pilgrim Aimery Picaud that <Urcia> was the word for `God' used by his Basque interlocutors. Note, however, that Picaud's is the *only* explicit testimony we have that this word was ever the name of a deity. Still, it *is* explicit, and the other terms recorded by Picaud seem to be very accurate.

None the less, there are problems. For one thing, though Christianity came late to the Basque Country, it was almost certainly the religion of the people by the time Picaud came on the scene in the twelfth century, and hence it would seem surprising that the Basques should hav been producing the name of a pre-Christian sky god when asked for their divine name. For another, Picaud records <Urcia> with the article, which is normal enough with a common noun but rather unexpected with a proper name.

There is perhaps a simple explanation available. Michelena (1964b: 51) suggests that, when Picaud asked for the name, he pointed to the heavens to try to clarify what he was asking for, and that the Basques, misunderstanding his intention, simply gave him their word for `sky'. After all, <ortzi> is well attested in the meaning `sky'.

This, of course, doesn't explain why the word <ortzi> should be present in the name of Thursday, but Michelena has another suggestion, somewhat unexpected. He proposes that the name of Thursday doesn't contain <ortzi> at all; instead, it is simply a compound from <bortz> ~ <bost> `five': hence *<bortz-egun> `fifth day'. The loss of initial <b-> before <o>, recall, is so frequent in Basque as to be practically regular, though the word for `five' itself happens to be the outstanding exception.

Against Michelena's proposal is the observation that no other day-name has such a formation. Moreover, we are left with the problem of the name of Friday: <ortzirale> and variants. This seems to contain the same first element plus an unidentified second element.

B <eguen> `Thursday' (variants <eguun> and <eguaun>) appears to be from *<egun-egun>, curiously `day-day'. B <bari(a)ku> `Friday' is confidently derived by Michelena from an original *<abari-ba(gari)ko-egun> `day without dinner': Friday is, after all, a fast day in Catholic countries. B <egubakoitz> is <egun> plus <bakoitz> `unique'.

The name of Saturday, <larunbat>, has attested variants <larenbat> and <laranbate>, and the Refranes y Sentencias of 1596 cites the form <lauren bat>. As first pointed out by Vinson, this looks very much like a reduction of <laurden bat> `one quarter': if Saturday is taken as completing the week, then each Saturday represents (approximately) one quarter of the month. Michelena, however, is skeptical: he is inclined to see this name as derived somehow from *<laur(en)-egun> `fourth day', since, in the old Roman style of counting, Saturday is the fourth day after Wednesday, the <asteazken> of Basque. He notes, however, that the <laranbate> of the sixteenth-century author Leizarraga may well preserve a final vowel lost elsewhere, in which case this name might involve the suffix <-te>, which possibly occurs also in <igande> `Sunday' (Z R <igante>), <aste> `week', <urte> `year' and <mende> `century (Z R <mente>).

The regional <egubakoitz> is again `unique day'; <ebiakoitz> is an alteration of this, as very likely is <irakoitz>.

Bizkaian <zapatu> is a loan from Latin <Sabbatu> `Sabbath', showing the interesting development of Latin <-bb-> to Basque <-p-> discussed in Chapter 3. The engaging eastern <neskanegun> is `girls' day'.

The name of Sunday, <igande> (Z R <igante>) is troublesome. Somewhat reluctantly, Michelena considers the obvious derivation from <igan> `ascend', since the Christian Ascension took place on a Sunday.

The Bizkaian <domeka> is a loan from Latin <(dies> dominica> `(day) of the Lord'.

REFERENCE: R. L. Trask, 1997, The History of Basque, London: Routledge, pp. 277-280

Larry Trask
University of Sussex
Brighton BN1 9QH