i "; $bpudate ="Sun, Jun 22, 2003"; $bpupath ="../../"; $bpuauthor ="Larry Trask"; $bpusubtitle=""; $bpuwhen =""; $bpunotes ="yes"; $bpuspecial="trask"; ?>

Since this list is reviving, and since somebody has expressed interest in information on the language, I thought I might start sending out the occasional posting on what we have learned about the prehistory of Basque.

I begin with an interesting suffix. It seems certain that Basque long ago had a suffix <-i> which formed adjectives. This suffix has not been used to form new words for many centuries, but its former presence is still detectable in the modern language.

The clearest case is this pair:

        <gatz> 'salt' > <gazi> 'salty'

The alternation here between <-tz> and <-z-> is familiar and well understood.

Almost as clear a case is this adjective:

        <garai> 'high'

Now, in spite of Azkue's somewhat misleading account, the required base word *<gara> is nowhere recorded in Basque in a relevant sense. Still, the former existence of this *<gara> seems certain. It must have meant something like 'high place' or 'elevation', and apparently it also had a transferred or metaphorical sense, roughly 'head' or perhaps 'skull'. This vanished *<gara> is still visible in several formations, such as <garaun> ~ <garun> 'brain(s)', whose second element is the variant <un> of the word which appears variously as <un> ~ <hun> ~ <mun> ~ <muin> ~ <fuin>, and so on, meaning 'pith', 'marrow', 'inner part' -- hence, literally 'head-marrow' or the like. It also forms the first element in the widespread word <garondo> 'neck' -- literally, 'head-bottom'. It also appears in the eastern words <garkola>, <garkhotxe> 'nape of the neck'; the second element in the first is obscure, but the other second element appears to be eastern <khotxe> 'basin', 'concavity', since the nape is concave.

This same *<gara> appears in several surnames, such as <Garate> and <Azkarate>. It may also be present in the place name <Bergara>, though this is less certain.

There are many other adjectives ending in <-i>, and we may be confident that at least some of these contain the suffix <-i>, even though the bases to which the suffix is attached have been lost as independent words and are therefore obscure. Here are a few examples:

        <berri>  'new'
        <bizi>   'alive'
        <eri>    'sick, ill'
        <gorri>  'red'
        <guzi>   'all'
        <hori>   'yellow'
        <itsusi> 'ugly'
        <larri>  'nervous'
        <lasai>  'loose, calm, relaxed'
        <sarri>  'frequent' 
                 (today usually an adverb, but formerly an adjective)
        <urduri> 'nervous'
        <zuri>   'white'

In a few cases we can make plausible guesses about the stems.

For example, <berri> 'new' is perhaps built on a stem *<berr->, which may also be present in <bertze> 'other', which is more conservative than its western variant <beste>. (Sorry; I can't remember who first suggested this.)

The word <gorri> 'red' may be built on a stem *<gorr->, which may also be present in <gordin> 'raw', since <-din> is another familiar adjective-forming suffix. Azkue suggests that this *<gorr-> might have meant something like 'flesh', which is plausible but beyond investigation.

The case of <guzi> 'all' is exceptionally interesting, since this eastern word corresponds to western <guzti>. As it happens, <-ti> is another familiar adjective-forming suffix, found for example in <bildurti> 'fearful' from <bildur> 'fear'. It may be, then, that a stem *<gu(t)z-> gave rise to two different adjectival formations, *<guz-ti> in the west and *<guz-i> in the east. (This is my own idea.)

For <hori> 'yellow', Azkue suggests as a source the old word <or> 'dog', found in early Bizkaian and still preserved in the east today. If Azkue is right, the original sense would have been something like 'tawny'.

For <zuri> 'white', Azkue suggests as a source <zur> 'wood', in which case the original sense would have been 'wood-colored', 'light-colored'.

Both of these last two suggestions are plausible but beyond investigation.

Of course, not all adjectives in <-i> can possibly contain this suffix. For example, <handi> 'big' and <garbi> 'clean' probably don't, since *<hand-> and *<garb-> look like impossible stems in Basque. And <tipi> ~ <tiki> 'small', with its diminutive forms <ttipi> and <txiki>, looks nothing like an ancient Basque word. This is surely an expressive formation comparable to Middle English <tine>, unrelated modern English 'teeny' and 'teensy', French <petit>, Spanish <chico>, and so on. Expressive formations for 'small' built on a syllable like <ti-> or <chi-> are exceedingly common in the world's languages.

Other non-candidates include <gori> 'fiery' and <busti> 'wet', both of which are borrowed from other languages.

Clearly we can't unravel everything. Still, we can safely reconstruct an ancient adjective-forming suffix <-i>, which long ago was probably used to form adjectives with some freedom, but which has long since ceased to be productive and given way to other suffixes, such as <-dun> and <-tsu>. Sadly, it seems impossible to determine the semantic value of this suffix, if indeed it ever had any.

Moving into the realm of pure speculation, we might wonder also whether certain nouns ending in <-i> might long ago have been adjectives -- for example, <harri> 'stone' and <herri> 'inhabited place'. It is not inconceivable that <harri> might once have meant 'stony', and that the word underwent a shift of meaning to 'stone'. But there is no earthly way we can investigate such speculations: the true story is lost in the darkness of time.

Finally, I note with amusement that English also has an adjective- forming suffix /-i/ -- spelled <-y> -- which is still productive today. A few examples of the very many formations:

        dust / dusty
        dirt / dirty
        fog / foggy
        mud / muddy
        ice / icy
        snow / snowy
        haze / hazy
        rubber / rubbery
        steel / steely
        trash / trashy
        glass /glassy
        foam / foamy
        sex / sexy
        star / starry

The English suffix descends from Old English <-ig> ~ <-aeg>, which descends in turn from Proto-Germanic *<-iga> ~ *<-aga>, which descends in turn from Proto-Indo-European (PIE) *<-i-ko-> ~ *<-a-ko->. The PIE suffix here is in fact *<-ko>, and the preceding vowels are merely stem vowels. In English, the original suffix *<-ko-> has been worn away completely, and all that is left is a remnant of the ancient stem-vowel.

Perhaps the Basque suffix has a comparably elaborate origin, but I'm afraid we will never know. It's just about all we can do to recognize the former existence of adjectival <-i> at all.

Larry Trask COGS University of Sussex Brighton BN1 9QH UK


Tel: 01273-678693 (from UK); +44-1273-678693 (from abroad) Fax: 01273-671320 (from UK); +44-1273-671320 (from abroad)