Introduction
These comments were posted to an Internet Discussion List called "Evolution of Language" (EvolutionLanguage@list.pitt.edu) during August and September of 1998 by Patrick C. Ryan in response to criticisms of the concept and content of the Basque essays entitled
This is not the first occasion on which I have discussed these matters with Professor Trask.
Therefore, I know that Professor Trask has many general objections to a possible relationship of Basque with any other language or language family on earth; and, I may have reason to comment on some of these in other supplementary essays.
But in this essay, I will principally restrict my comments to a strenuous objection of the use by Professor Trask of the term ‘expressive' to reject proposed cognates a priori by the simple device of his labeling them ‘expressive'.
In a posting to the "Evolution of Language" list dated 8/20/98, Professor Trask objects to the comparisons of Basque milikatu, ‘lick', and diti, ‘nipple, breast', in the following manner:
Having labeled them ‘expressive', Professor Trask believes the matter has been satisfactorily disposed; and he offers no additional information beyond the labeling to buttress his implied claim that these words may not be legitimately compared.
Now a look at a respected English dictionary like The American Heritage Dictionary of the English Language, which includes many terms from linguistic terminology, does not help us much in understanding Professor Trask's use of this term:
Professor Trask has written and had published A Dictionary of Grammatical Terms in Linguistics (1996. Routledge: London and New York) but it has no entry for 'expressive', which one might expect. Similarly, in his A Dictionary of Phonetics and Phonology (1996. Routledge: London and New York), it is also noticeably absent. Since, as we shall shortly see, his concept of 'expressive' seems to be connected with phonaesthetics, which does have an entry, this omission seems remarkable.
In order to get even a vague idea of what this term means for Professor Trask, we must turn to his The History of Basque (1997. Routledge: London and New York), where we find two page references in the Index under "expressive formations, 128, 257-259".
In page 128, we read: ". . . most words with initial m today are loan words or recent formations, often ‘expressive' in nature . . .", which does not elucidate the usage of ‘expressive' as favored by Professor Trask.
On pages 257-259, we can perhaps gain a little idea of what Professor Trask has in mind when he uses ‘expressive'. Contained under a heading "5.5 Sound Symbolism", Professor Trask proceeds to list a number of word categories:
a. If Professor Trask should suggest that ‘nursery words' in various languages have no intrinsic relationship but are independently re-invented anew by children in many, many, cultures, I would have to respond that if children in many cultures re-invented the phonologically same word anew for ‘excrement' independently, then this circumstance would likely be an indication that the earliest language of man must have had it also. In order to believe that it subsequently had to be re-invented in the languages in which it is present, and not derived from any earlier language, we would have to assume that adults forget the words of their childhood so that children can re-invent them in exactly the same form. In my opinion, this is plainly ridiculous!
b. Now, let us ask ourselves why this patently foolish ‘nursery word' objection has enjoyed the currency that it admittedly has. If one assumes that Language A and Language B are not genetically related, and both languages have a word for ‘excrement' that phonologically resembles kaka, then the only possible conclusion will be that some process which is non-genetic must account for the ‘coincidence'; and ‘nursery words' being re-invented by children could be a superficially good guess but its superficiality would be revealed when we ask ourselves why every language does not have a similar ‘nursery word' for ‘excrement'. If there were something ‘excremental' ("phonaesethetic"ally speaking) in this sequence of sounds (kaka), then every language should have it as a ‘nursery word' for ‘excrement'. But every language does not! Therefore, since some children do not ‘invent' it, it is not likely that any children have ‘invented it'.
On the other hand, if we allow that Language A and Language B could be genetically related, then it would be natural and expected to find that ‘nursery words' which were cognate might be present. The most plausible exception to this surmise would be that the circumstances could be traced to borrowing but borrowing among children of the nursery does not seem very likely at all.
c. Since beginning these discussions with Professor Trask many months ago, I have carefully listened when I go into my bathroom; and I can verify from personal observation that kaka does not appear to be ‘onomatopoeic' either.
d. Therefore, for all of the above reasons and others, I totally reject Professor Trask's claim that some Basque nursery words (e.g. kaka and diti) may not legitimately be compared with words in other language families provided that regular correspondences among other ‘nursery words' and ‘non-‘nursery words' can be demonstrated to exist.
3. Onomatopoeic Words. Now, with a word like B dzast!, which Professor Trask renders as ‘bang!', this may well be a purely Basque coinage since I do not believe that a speaker of any language other than Basque would have the slightest idea what was meant if he heard dzast!. I also have never heard an explosion, to imitate the sound of which I would attempt to say anything vaguely resembling dzast!. But with a term like murmur, ‘murmur, whisper', I am not at all certain that we can be so cavalier. In the form *mormor- / murmur-, this term is attested in Pokorny for twelve IE-derived languages. Now if Professor Trask tells us this is because of a universal tendency to hear murmuring as **mVr-mVr, one can legitimately ask, why the probable equivalent marmara in Arabic (which, by the way, exists in Basque also as marmar, ‘murmur') is restricted to ‘grumble'? Cognate with German murren, ‘grumble'. And why, if this is a purely onomatopoeic word based on an imitation of the sound that accompanies grumbling, there is no equivalent expression in English?
In Egyptian, an Afrasian language related to Arabic and through Nostratic to Indo-European, we find no trace of a mrmr or a m3m3; presumably the Egyptians murmured and grumbled like the rest of us. Why is it not there?
In fact, does not **mVr-mVr behave exactly like non-onomatopoeic words in that it is retained in some languages, discarded (or never created by imitation[?]) in others, and undergoes normal semantic restriction or shifts in yet others?
Therefore, it is irrational to consider a Basque word "onomatopoeic" because it is reduplicated, or has a phonological shape that Professor Trask does not want to admit could exist. Is zurrut, ‘gulp', onomatopoeic as Professor Trask would have it, or could it possible be related to LN, U zurrupa, ‘slurping'. Perhaps Professor Trask would say that zurrupa is also onomatopoeic, in which case we would be debarred from considering a connection with IE srebh-, ‘slurp up, drink noisily'.
And, of course, the final recourse, when all else fails, is to call a word like zurrupa a loan from a neighboring language, even when the word is unattested in that language. This, of course, might mean something in some contexts, but I am arguing that every language in Europe (at a minimum) is related through common descent, so I would hope to see reflexes of any given early word in any of them. If the form of the Basque word conforms to what is expected from the tables of correspondence, then I can safely discount the possibility that it has been borrowed from any other language.
Since Professor Trask stoutly maintains that /m/ did not exist in early Basque, a word like murmur is also ruled out as a comparandum because it simply could not exist. For ‘murmur' or ‘mumble', I guess we have to assume that Basques, in distinction to the IE and AA peoples around them, said something like **burbur, which, strangely, has not survived in any dialect known to my dictionary. Therefore, Professor Trask will tell us that murmur has to be a recently coined word because it could not have existed at an early date; and that there is no early Basque /m/ because words that have it now, and might be thought to have possibly had it then, are all ‘expressive', and cannot be counted as normal words.
Is it any wonder that I am certain that Professor Trask's favorite pastime as a child must have been a merry-go-round? This is another patently circular argument.
Therefore, in summary, let me state that I believe Professor Trask illegitimately uses this classification ("onomatopoeic") to debar comparisons that might otherwise conform to established procedures for establishing cognation.
4. Reduplicated "but not strictly onomatopoeic" Words. An example of this genre is zurru-zurru, ‘in gulps'. I am not sure why Professor Trask would count zurrut, ‘gulp', as onomatopoeic (which, by the way, it is not to me), and zurru-zurru as "not strictly onomatopoeic" (because it is reduplicated?). But, because he has classified it so, "expressive", that is, we are prohibited from comparing it with words like Egyptian zwr (/s(h)awar/), ‘drink', or Arabic saraTa, ‘gulp' (remember zurrut?), or Indo-European **swer-, ‘gulp', in Old Indian su:ra, ‘intoxicating drink'.
I think it is obvious that I feel that classification has been principally invented to debar comparison of Basque words with words in other languages when no other plausible prevention technique is available.
5. Adjectives in initial m. Here Professor Trask includes a number of nouns and adjectives "denoting physical or moral defects". An example would be matzer, ‘deformed, twisted, defective'.
We have been told on page 128, that "Probably no word beginning with m existed in Pre-Basque with this initial, apart from one or two loans from Celtic, such as mando ‘mule'." The question that might occur to some linguists reading this is how Basque, possessing no /m/ would easily borrow from Celtic a word with initial /m/? This is not the pattern we usually see, is it? Is it not very common for a sound that is present in the borrowing language to be substituted for the sound in the word of the donor langugae that that is not present in the borrowing language?
In any case, I am sure some readers like myself will find this scenario just a little unbelievable. But, to return to the classification mentioned above, then Professor Trask proceeds to tell us that these one or two of these nouns and adjectives ". . . may possibly be derived with the ancient ‘expressive' prefix ma- from other words like oker ‘twisted'".
Now, evidently, the view expressed on page 128 should be emended to read: there is no m in early Basque unless it is the initial of a word that Professor Trask considers to be ‘expressive', in which case it is permitted. So we have a language, according to him, that prohibits all m except in loanwords and ‘expressive' words. I cannot think of any other language with which I am familiar which has one sound-system for normal native words, and another sound-system for foreign loanwords and ‘expressive' words, whatever that might mean. On that basis, perhaps I will be forgiven for characterizing in my own mind this view by one of the ‘nursery words' we discussed rather exhaustively above.
6. Words with initial m which Professor Trask has not been able to fit into any other classification.
7. Phonaesthetic Words. Professor Trask explains; "By a ‘phonaesthetic' word I mean one which has apparently been coined out of thin air purely because of its appealing sound".
This is not the definition given in Professor Trask's A Dictionary of Phonetics and Phonology, where we read instead: "phonaesthesia . . . Any direct correspondence between sound and meaning . . . and such cases as Basque tximeleta ‘butterfly', whose delicate, fluttery sound appears to mimic the appearance of the insect".
To me, tximeleta suggests the sound of a fragile argument that will not break into a thousand pieces only if it is not looked at closely.
As representative of this classification, we might notice karramarro, ‘crab'. I am not sure how Professor Trask would explain what sound karramarro mimics that could pertain to the appearance of the crab but frankly, I think it is more than coincidence that Indo-European *3. kar- is the basis for words in the derived languages meaning ‘crab', and the sea in which crabs reside, is frequently connected with the Indo-European root *mori-, ‘sea', but, of course, Professor Trask will assure us that this is only a coincidence; and I guess, I should ask him, if he considers Old Indian karka-H, ‘crab', an analogous phonaesthetic creation (of recent times?) to karramarro.
We might mention, before we leave this ‘expressive' list that has suppressed much more than it has expressed, that included in this final classification is zurrumurru, ‘murmuring', which in view of "onomatopoeic" murmur, mentioned above, and zurru(nga), ‘snoring', is probably just a simple combination of the element seen in Indo-European *2. swer-, ‘hum, whir', and the element seen in mur(mur)- discussed earlier.
I trust this is rather obvious to most readers of this essay so why has this not occurred to Professor Trask to whom I nor anyone else should impute stupidity?
It may seem that I do not hold Professor Trask in high regard — but I do. Without his excellent The History of Basque, we would be forced to learn Spanish, a language of limited scholarly usefulness, in order to obtain the information contained in it.
But I honestly believe that Professor Trask (and possibly a limited number of other Vasconists) love the Basque language so arduously that they cannot bear to share it through affiliation with another language family.
In pursuit of this passion, a number of theories have been devised to keep Basque in grand isolation; and inspite of the inadequacy of the arguments supporting this false view, they have been sufficient to keep the wolves at bay because of the relatively inconvenient accessibility of Basque grammatical materials and rather limited scholarly interest in Basque.
So, I attribute to Professor Trask only a crime passionel