Re: I've recently begun studying Turkish, and one linguistic feature has struck me in particular
--- In email@example.com, Polat Kaya
Turkish is also unique in the sense that both the nagative and the
positive of a verb is done by the same "-ma, -me" suffix. for example:
"gelme" meaning "dont come", "gelme" meaning the act of "coming";
"okuma" meaning "dont read", but "okuma" meaning act of "reading";
"konusma"/"konusma", etc.. Thus in one case it is negative, in the
other it is "positive". This is also one of the most unique feature
of the Turkish language.
The "MAYA" (South American) language also uses suffix "-ma" for
negation. However, I am not sure whether it uses this suffix in the
same way as Turkish does. Some of you may be able to enlighten this
aspect of Maya language. There seems to b many similarities between
Maya and Turkish languages. I believe they had some comonality among
them in the past.
Kamil KARTAL wrote:
> Kimden:gec (geoffreycaveney@y...)
> Konu:Turkic verbal negation: unique among world's languages?
> View: Complete Thread (20 mesaj)
> Original Format
> Haber Grupları:sci.lang
> Tarih:2003-05-13 03:53:38 PST
> I've recently begun studying Turkish, and one linguistic feature has
> struck me in particular: the basic means of expressing negation of a
> verb. As you probably know, in classic agglutinative style Turkish
> expresses negation with the infix -ma-/-me- (the choice dependent on
> back/front vowel harmony). It follows immediately after the verb
> and before any and all conjugational markers denoting mood, tense,
> person, number, etc. Thus it falls smack dab in the middle of
> well-known examples of super-long agglutinated words such as
> 'AvrupalIla$tIrmadIklarImIzdansInIz' = "You are one of those whom we
> did not Europeanize" -- which one must be careful to distinguish
> 'AvrupalIla$tIrdIklarImIzdansInIz', which means the exact opposite.
> [Transliteration note: $ = "sh"; I = the high back unrounded vowel,
> written in Turkish as an "i" without a dot.]
> That's an extreme example, but the following one is not: some of the
> most basic essential sentences any beginning language learner needs
> know are the ones that mean "I know", "I don't know", "I understand,
> "I don't understand." In Turkish those would be, respectively,
> 'biliyorum', 'bilmiyorum', 'anlIyorum', 'anlamIyorum.' [Before the
> present tense -yor- suffix, the vowels a/e reduce to I/i, and as
> as I can discern that syllable tends to be swallowed up in normal
> speech, sounding like bilyorum, bilmyorum, etc.] In fact this is
> the Turkish negation feature really struck me: How many languages
> there in the world in which the vital distinction between the
> and negative forms of such essential units of meaning is expressed
> as phonetically unemphatic a morpheme as [m] wedged between [l] and
> [y] in the middle of a word?
> I began looking for any similar examples from non-Turkic languages.
> Agglutination itself is a rather common language characteristic,
> came across general comments which implied that a negative infix
> normal feature as well: "A language's favorite way to express
> sentential negation may be an inherently negative verb, such as
> Finnish ei, or a negative infix in a morphologically complex verb,
> such as Turkish ma, or a negative adverb or particle, such as
> classical Greek uh." [See
> <http://odur.let.rug.nl/~vdwouden/docs/zanutt.ps>, or Google's
> version thereof, first paragraph.]
> But so far, I've yet to find a non-Turkic example of a
> infix in a morphologically complex verb."
> (1) The prime candidates would of course be the agglutinative
> languages once conjectured to be related to Turkish as part of an
> "Altaic" language family: Hungarian, Finnish, Mongolian, Korean,
> Japanese, etc. None seem to fit the bill:
> Hungarian uses a good old-fashioned Indo-European style negative
> particle, 'nem': "I [don't] know/understand" = 'Tudom', 'Nem tudom',
> 'Ertem', 'Nem ertem.'
> Finnish uses the above-cited negative auxiliary verb 'ei', which
> conjugates: "I don't read" = 'En lue'; "you don't read" = 'Et lue.'
> Another Finnic example is in Sami: "I don't hear" = 'In gula' as
> opposed to 'Gulan' = "I hear."
> Mongolian uses a suffix: "I don't understand" = 'bi oilgokhgwee
> vs. 'bi oilgoj bain' = "I understand." For me, this fundamentally
> differs from the Turkish negative in two key respects: (a) the
> negative suffix is word-final, and (b) it is much more phonetically
> prominent than the Turkish [m] infix.
> Japanese also uses prominent, word-final (as far as I understand the
> examples) suffixes for verbal negation: 'kaka' = "write", 'kakanai'
> "not write"; 'os' = "push", 'osanai' = "not push"; 'tabe' = "eat",
> 'tabenakatta' = "not eat".
> Korean may use either a verbal prefix or a separate particle in the
> middle of the verb phrase to express negation, but not an
> agglutinative infix: 'Johni ppangul mekessta' ("John ate the bread")
> can be negated as 'Johni ppangul anmekessta' or 'Johni ppangul mekci
> ani haessta' ('ani' being the negative particle), but not, say,
> *'Johni ppangul mekanessta'. Note also that in the particle form,
> morphemes before and after the particle are modified as well ('-ci'
> and 'ha-'), further marking the negation.
> (2) My perusal of a wide variety of other languages from all around
> the world has so far turned up nothing that resembles the Turkish
> system of verbal negation:
> The overwhelming preference seems to be for an unchanging particle
> particles to mark negation. Thus one finds Nahuatl 'amo',
> ... da', Arabic 'mish', and numerous others that I didn't mark down
> because I knew I wasn't looking for languages with negative
> Negative affixes are either word-final like in Mongolian, or
> word-initial like in Korean, or Persian: 'miravad'-'nemiravad' ("he
> going"-"he is not going"), 'rafte ast'-'narafte ast' ("he goes"-"he
> does not go").
> Quechua combines a phrase-initial particle 'mana' with a later
> word-final suffix '-chu'.
> Related to the Finnic negative auxiliary verbs is the Dravidian
> so-called "negative conjugation", e.g. Kanarese 'mOtlidenu',
> 'mOtlidevu' ("I did", "we did"), etc. vs. 'mOdenu', 'madevu' ("I did
> not", we did not"), etc.
> [The source of these forms was confusing, and none too generous with
> explanations, so I may not have them exactly right. But it looks
> a synthetic negative conjugation, and not some sort of agglutinative
> negative infix.]
> (3) I have had trouble locating on the Internet examples of
> positive/negative contrastive forms in African languages, so in many
> cases I am left with the brief descriptions of different languages'
> negation systems, made in passing in papers focusing on other
> But in the languages for which I can find concrete data online --
> Swahili, Wolof, and a bit of Yoruba -- nothing looks like the
> Swahili has a negative conjugation which appears as a combination
> set of word-initial prefixes with a change in the word-final suffix.
> Thus 'najua'-'sijui' ("i know"-"i don't know"), 'wajua'-'hujui'
> know"-"you don't know"), 'ajua'-hajui'; 'twajua'-'hatujui',
> 'mwajua'-'hamjui', 'wajua'-'hawajui'. My personal aesthetic reaction
> is that this system is quite elegant. But again, despite certain
> characteristics of agglutination, it fits the prevailing tendency
> any negative verbal affixes are word-initial or word-final.
> Wolof, somewhat similarly though more simply, has a set of word-
> suffixes which may be said to constitute a negative conjugation.
> Interestingly, in positive forms the modifiers precede the root-verb
> morpheme as separate words, while the basic negative form is a
> which follows it. Thus 'Maa ngi wax'-'waxuma' ("I am speaking"-"I am
> not speaking").
> For Yoruba, the only relevant phrase I could locate was 'ko ti i ye
> o' = "I don't understand", and the fact that 'ko' is the negative
> element. So all I have to go on is that in one very important basic
> vocabulary item, the negative marker is phrase-initial.
> (4) I would love to know if there are any non-Turkic languages in
> world with an agglutinative negative infix. I think the need for
> clearly-marked negation militates against it, but if it could arise
> Turkish there's no reason it couldn't have arisen somewhere else. In
> fact I have a hypothesis about how the situation arose in Turkish:
> Turkish has an aorist as well as a present tense. My introductory
> on Turkish explains that in modern colloquial usage the present
> is replacing the aorist "quite systematically." That is,
> the aorist was used for "timeless" statements of fact, while the
> present was used to describe actions actually taking place at the
> present moment. Now the present is coming to be used for both those
> types of statements, while the aorist's main productive usage is for
> statements of future intent; that is my understanding for example of
> the morpho-syntax of a common phrase when saying
> 'gör-' is the root-verb "see", '-ü$-' is the reciprocal verb suffix
> ["each other"], '-ür-' is the aorist tense marker (future intent),
> '-üz' is the personal ending "we". Thus, all together, "We'll be
> seeing each other", i.e., "See ya."
> The significance of this semantic shift for Turkish negation is that
> the negation of aorist forms is much more clearly marked than the
> negation of the present tense forms. The aorist for "I understand"-
> don't understand" is 'anlarIm'-'anlamam'; for "I know"-"I don't
> 'bilirim'-'bilmem'. The same holds for the rest of the conjugation:
> 'bilirsin'-'bilmezsin', 'bilir'-'bilmez', etc.
> The aorist negative forms in '-mem' and '-mez' are quite exceptional
> in Turkish grammar. Everywhere else the negative infix '-me-/-ma-'
> always followed by the same regular tense markers which appear in
> positive forms; here the aorist '-r-' disappears, replaced by '-z-'
> the 2nd/3rd person and by nothing in the 1st person. Also, G.L.
> Lewis's 1967 _Turkish Grammar_ (which incidentally may have predated
> the aorist-->present usage shift, as it doesn't appear to mention
> notes that in the aorist the negative syllable itself is stressed,
> whereas elsewhere the stress is on the syllable preceding the
> My hypothesis is that these exceptional features were connected to
> fact that basic general statements of affirmation and negation used
> be made with this verb tense in Turkish. Hence negation was
> with additional modified elements beyond the expected agglutinative
> negative affix itself. Compare the Korean example above where
> 'mekessta' becomes negative 'mekci ani haessta' and not
> Likewise in the Turkish aorist, where 'bilirim' becomes 'bil-MEM'
> not *'BIL-merim'. Note further that the negative syllable ends up in
> word-final position in the 1st and 3rd person singular forms.
> The forms thus fit their semantic functions well -- until the
> tense began to supplant the aorist as the form used to express
> statements of fact. Now a linguistically anomalous situation is
> arising in Turkish where negation is not well marked. The above-
> review (van der Wouden, of Zanuttini) references Jespersen's work on
> the instability and cyclic tendency of negation systems. I wonder if
> modern Turkish is approaching a point on a "Jespersen's cycle" where
> it is ripe for one of its non-verbal sentence negation markers
> = "there isn't"; 'deyil-' [spelled 'degil' with soft-g] = "...is
> not..." or "it is not that...") to begin to be used as a
> negation marker in verbal sentences?
> This hypothesis would predict the appearance of such forms as
> *'bilmiyor deyilim' ("I don't know");
> *'anlamIyor deyilim' ("I don't understand")
> That would produce the interesting agglutinative verb-phrase syntax
> [root]-[Neg1]-[mood/tense/etc.] [Neg2]-[person/number].