Duals and Comparatives:An e-mail conversation

This is an e-mail conversation that I had with :

Prof.Gautam Sengupta (GSG), Probal Dasgupta (PD), Prof. Foster

Chinmay to GSG & PD:

Sir,

I am trying to see some relation between

the duals and the existence of Comparative Degree form

in Indo-European languages.What is interesting is the fact that

though the modern Indo-European languages have lost the duals

some of them retain the Comparative forms e.g English,German etc.

The fact is true with the Indo-European alone.Dravidian Family never

had duals and nor the comparative forms.No wonder in that.

However Semitic languages had and still retain the dual forms

and never had the comparative forms morphologically independently formed.

So the Comparison is analytical as in Hindi,Marathi (save the very Sanskritised
bookish forms rarely used: sundartara,sundartama etc)

So the relation between the duals and the comparative forms is

idiosyncratic to the Indo-European languages.

What I want to know form you is that if there is any other family

- languages of which have the duals and the comparative forms?

Chinmay.

GSG to me:

Dear Chinmay,

The facts, and the evidently non-random correlation between them seem
very interesting. This line of research may have a lot of promise. Try
posting your query on the Linguist List, so that linguists speaking
diverse languages get to see and respond to it. I myself do not have
that kind of encyclopediac knowledge.


Best, G.

Then on Gsg's suggestion I posted my doubt on linguist's list:

and what I got is as follows:

Prof.Foster to me:

Found it have I that your query has stayed unanswered. I was hoping that a panelist who knows more than I do about this might reply, but I will take a shot at it.

Speak for my colleagues I can not, but one of my difficulties is trying to determine what you mean by "some relation between the duals and the comparative degree forms [of adjectives / adverbs?] in Indoeuropean and then your statement "This fact is true with Indo-European alone...".

I am not certain which "fact" you refer to. Yes, *Proto Indoeuropean apparently had a grammatical dual in the nouns -- or at least some nouns. And most Indoeuropean languages have lost it, although there are more traces around of it than you might be aware of. Welsh and the Gaelic still have traces of it--though differing in details; Slovenian still has traces of it. And English of course has either ~ any and both ~ and -- "??All two of them came. instead of Both of them came. sounds odd. And for old curmudgeons like me, a sharp distinction of between versus among is still very alive.

But while the grammatical number 'dual' is relatively uncommon among languages of the world, it is widespread and found in a number of language families other than Indoeuropean.

Now for the comparison of adjectives. I think what you are referring to with "the comparative form morphologically independently formed" is the formation of comparison of adjectives with a special affix, in this case a suffix, on the positive form, thus as in English far, further (or farther), furthest, furthermost with a suffix -(th)er, (th)est, and the absolute superlative like the superlative but with the word 'most', here an enclitic, added on. German: fern, ferner, fernest, am fernsten. Italian, bravissimo "super bravo!" and Italian can even use the absolute superlative on nouns -- Generalissimo. 'Supreme General'. So can Hungarian, to which we will return below.

Yes, this use of a special comparison suffix to mark the predicate adjective was a characteristic of Indoeuropean, and still retained in Slavic, Germanic, and Celtic, lost in most of the Romance languages except for traces in, usually suppletive forms, like Spanish buen ~ mejor 'good~better' where the comparative has an -jor, a reflex of the Latin -ior, the usual comparative ending (bon ~ melior).

Now, so far as I am aware, it is accurate to say that this comparison affix on a predicate adjective is very rare in languages and is primarily -- possibly exclusively now a phenomenon found in European languages. But it is not quite accurate to claim that is "found only in Indoeuropean". Basque, certainly a European language but equally certainly not an Indoeuropean one, has four degrees of comparison, positive, comparative, superlative, and excessive, all but the positive shown with a suffix. Here is the paradigm for the adjective hotz 'cold'

hotz, hotzage, hotzena, hotzege 'too cold'.

Hungarian, a FennoUgric language, uses a vowel harmonic -abb/ebb suffix on adjectives -- and nouns -- to form a comparative. Thus

Janos - NOM magas-abb mint Istvan-NOM. 'Janos tall-er than Istvan.'

An alternative form uses the harmonic suffix -nal/-nel on the noun that refers to the standard, as in

Janos magasabb Istvan-nal..

This suffix is the adessive case and means 'nigh unto, close to', so what the latter (there's a comparative with the -er suffix) sentence means literally is 'Janos is tall-er near to Istvan.'

Hungarian can also derive a comparative adjective from a noun with the -abb/ebb suffix. The noun meaning 'fox' is roka. The sentence Janos rokabb Istvanal. means literally 'Janos fox-er Istvan-next to [is].', i.e. 'Janos is foxier than Istvan.'

Now, I believe that comparative Uralicists have evidence to reconstruct this kind of system for *ProtoFinnoUgric, and possibly for *ProtoUralic. So it may have once been more widespread.

So, what is rare is the suffixation of an adjective to show degrees of comparison. The presence of a grammatical dual number is not common but not exceptionally rare. Statistically, the two phenomona do not appear to be associated. But you seem to think that dual number has something to do with morphological affixation for comparison of adjectives in Indoeuropean. I cannot think of much in the way of particular reasons to posit such an association except for one possible case. I wonder if you might be thinking of forms like English 'whether', Latin 'utero', English 'other', Latin 'alter, ulterior, ' . These envolve a root and an old -tero, -toro, and in zero grade -tro suffix. The sense seemed to be something on the order of 'the one beyond' and came to mean 'the opposite of, the alternative person or item'. Now these do have a "pair' or "dual" sense. The dual suffix for *ProtoIndoeuropean seems to be a bit difficult of reconstruction --I think the most often given suggestion is -a, as in Slovenian volkova '2 wolves, a pair of wolves' versus singular and plural volk ~ volkovi, respectively. But I am not aware of any Indoeuropeanist's having proposed that the -tero suffix was or was related to the dual suffix, or suffixes.

But the -tero suffix meaning 'opposite, alternate' is etymologically related to the comparatives of adjectives in some IE languages. In Greek and Indoiranian it came to mark the comparative degree, and in Irish Gaelic the equative degree, as in luath 'swift' but luathither 'as swift'. This suffix may be cognate with the -ed suffix in Welsh used for the equative degree, as in gwyn, gwynned, gwynnach, gwynnaf 'white, as white, whiter, whitest', but I do not know that it is.

Soo.... I am not sure I have understood what you are asking, but I do not see either a general connexion in languages of the world between a dual number and morphologically marked comparison of adjectives, nor am I aware of evidence for it particularly in Indoeuropean. But it is an interesting idea and suggestion.

U of Cincinnati

Department of Anthropology

- - - - -

Joseph F Foster, Ph D

Associate Professor of Anthropology,

University of Cincinnati, Ohio, USA

Chinmay to Foster:

Sir,

thank you so much for your edifying reply.

what I mean by:

"some relation between the duals and the comparative degree forms"

is:

Note that comparative form is nothing but a way to compare between two.

Why such special treatment by giving morphological (against analytical) suffixation in case of adjectives(Sanskrit:sundaratara,uttara,uttungatara etc.) or pronouns (anyatara,itara etc) for comparison between two?

Interestingly and obviously enough, such forms are found in the

languages-- Pro-to of which had duals....do u have any such example...when pro-to doesn't have duals but the modern developments of which have separate comparative form?( If so it is exceptionally interesting?)This is precisely what I mean by 'some relation'.To rephrase in other words:

Existence of morphologically affixed comparative forms for adjectives has direct relation with the existence of dual number for nouns.

Also if this is not jabbering of mine:

With this as a point of departure we can trace the evolution of

counting numbers from 1,2 and their traces left in existence of duals and/or comparative forms.
Kindly comment to this mail.

Thank you once again for your the reply to the query.

Chinmay Dharurkar,

M.A.Applied Linguistics

Centre for Applied Linguistics and Translation Studies,

Hyderabad Central University,

Hyderabad 500046,

India.

Foster to Chinmay:

Dear Mr. Dharurkar,
I'm still not sure I am following your questions or proposals, but I'll try. Comments are inserted below after what they are commenting on.

At 12:51 PM 1/23/2007, you wrote:

Sir,
thank you so much for your edifying reply.
what I mean by:
"some relation between the duals and the comparative degree forms"
is:
Note that comparative form is nothing but a way to compare between two.

But actually, in many languages the comparative and superlative are not formally distinguished in any way.
And what about other degrees of comparison, excessive, insufficientive, equative, superlative, absolutive ( e.g. outermost in English.).
And even so, that does not mean there has to be a morphological or syntactic relationship, either synchronically or diachronically (historically) between the two. Dual marks a noun. Comparision of adjectives is a rather different device.

Why such special treatment by giving morphological (against analytical) suffixation in case of adjectives(Sanskrit:sundaratara,uttara,uttungatara etc.) or pronouns (anyatara,itara etc) for comparison between two?

Well, that's like asking why Khoisan languages have clicks since others don't or why a few languages have VOS order when most others don't. In general, we don't know. Some things that are rare nonetheless do exist. We would like to know why but usually it is difficult or impossible to find out why. And remember that in science, why really means "how it came to be". But as I noted, morphological suffixation to form the comparative and superlative, and equative (Celtic) or excessive (Basque) is not confined to Indoeuropean languages. Hungarian and some other Ugric languages have it, Basque has it. It does seem to be confined to Europe and Western Siberia.

Interestingly and obviously enough, such forms are found in the
languages-- Pro-to of which had duals....do u have any such example...when pro-to doesn't have duals but the modern developments of which have separate comparative form?

No, I dont happen to have any such examples. But that does not mean that dual number and suffixed forms of adjectival comparison are related phenomena. Dual number is fairly widespread. A number of languages in several different language families have it or have had it. But special suffixed forms of adjectives for degrees of comparison is far less common than dual grammatical number. (Incidentally, I don't think Basque has a dual number, but whether it did in the past I dont know and doubt if anybody does.) So having dual number is certainly not a sufficient cause for a language to develop suffixed degrees of comparison. Proto Uralic and Proto Indoeuropean both apparently had dual number. But it has been lost in Modern Hungarian but the suffixation of adjectives is very much alive. The same is true for Indoeuropean languages -- the dual number is much more restricted and vestigial but in the languages that have a suffix showing degrees of adjectival comparison, it is very much alive and productive (English, Russian, German &c. ) So one is very suspicious of positing a relationship between dual number, much of which once was, and suffixed comparison of adjectives, which very much still is. One is particularly hesitant to posit such a relationship when the sample size of languages with suffixed comparison morpheme is so very small and geographically proximate. It's just not a big enough sample to build much of a theory on.

( If so it is exceptionally interesting?)This is precisely what I mean by 'some relation'.To rephrase in other words:
Existence of morphologically affixed comparative forms for adjectives has direct relation with the existence of dual number for nouns.

But you have to have evidence that there in fact is such a relationship. You dont have any such evidence that I can see. Or,if you have, it is very weak. You have two language families -- Indoeuropean and Uralic -- in which there was a grammatical dual and in which there is found morphological marking of degrees of comparison of adjectives -- not just the comparative degree, but other degrees too. But there are many languages in which grammatical dual is found. The only other language in which there are suffixes showing adjectival degrees of comparison that I know of is Basque. That's not enough of a sample rule out coincidence as a likely explanation.

Also if this is not jabbering of mine:
With this as a point of departure we can trace the evolution of
counting numbers from 1,2 and their traces left in existence of duals and/or comparative forms.

I'm sorry, I don't follow this at all. We can find out about the development of words for numerals in lots and lots of languages. The occurrence of particular forms of nouns with particular numeral words may be an indication that there was an old partitive in a previous stage of the language, an old dual, or some other phenomenon.

And keep in mind that a few languages have a grammatical trial or paucal (a few) number in addition to a dual and a plural.

Joseph F Foster

Link to submit a question to Ask A Linguist on linguist list:
http://linguist.emich.edu/ask-ling/index.html