Talk:Navajo language

noun class for "long, stiff object" is a joke?

That noun class for "long, stiff object" is a joke, isn't it?

While I understand how you might think that, this is no joke. As I understand it, this noun class refers to things like sticks, rifles and planks, whereas long and non-stiff (i.e. flexible) things, like ropes, fall under a different noun class. Navajo has a large number of noun classes, and these correspond to physically observable characteristics of the noun in question. The assignment into noun class is not so arbitrary as in Latin or Greek. Here are some examples of a verb which varies according to noun class:

• ch'í-n-lhtí~ carry an animate entity out horizontally
• ch'í-n-lá carry a slender flexible object out horizontally
• ch'í-n-tá~ carry a slender stiff object out horizontally
• ch'í-n-jaa' carry many objects out horizontally
• ch'í-ni-'á~ carry a solid compact entity out horizontally
• ch'í-n-lhjool carry non-compact matter out horizontally
• ch'í-n-yí carry a burden or pack out horizontally
• ch'í-n-lhtsooz carry a flat flexible object out horizontally
• ch'í-n-ka~ carry something in a container out horizontally

thefamouseccles 10:08, 6 Sep 2004 (UTC)

That's right, this is no joke. Welcome to linguistics! I added a section about this. Check it out. - Ish ishwar 11:14, 2005 Jan 10 (UTC)

Polysynthetic??

For this reason, some call it a polysynthetic language.

I thought polysynthesis was the extreme use of affixes, not the contraction of affixes into one another: that's inflection.

Navajo uses up to 7 prefixes on its verb. I don't know what your definition of extreme is, but to me that's a lot. Also, the affixes don't just contract, sometimes their form changes depending on the segments around it (e.g. d-effect).

Of course this is all very subjective. I'll try to do some research into this and pull out some references from published papers, when I get some time. Until then I'll leave it out. - wulong 08:14, 8 Dec 2004 (UTC)

Some more thoughts.

1) Inflection is not contraction of affixes, but something quite different. I have put forth some effect to correct the inflection entry, but it still needs work. Check out the explanation there (or better yet consult an introductory linguistic textbook!). But at any rate, Navajo has quite a bit of inflectional processes going on.

2) Most Athabaskanists (i.e. linguists who work on Athabaskan languages like Navajo & Tlingit, etc.) consider Athabaskan languages to be polysynthetic.

But, if you check out what polysynthesis means you will see that what is being talked about here is the degree of synthesis or the word-to-morpheme ratio (i.e. how many morphemes are present in each word on average). So if we look at a bunch of langs we see that there is a continuum from langs that words with only 1 morphemes to ones that have 2-4 morphemes per word to langs that have quite a high number of morphemes per word.

Another thing to think about is "Do different word classes have different degrees of synthesis?". The answer is "yes". Take Japanese: the verbs generally have a greater degree of synthesis than nouns. Things are getting more & more complicated...

So where is Navajo? Well, it definitely has more morphemes in verbs than in nouns (the non-deverbal nouns) or other word classes. So maybe the verbs are polysynthetic but the nouns are just synthetic. But if we compare Navajo to some really polysynthetic languages like Chukchi or Central Siberian Yupik (a.k.a. Eskimo), we see that Navajo is not as extreme. So it depends on your definition of polysynthesis which is what Wulong stated above.

The other thing to mention is that these categories of isolating, synthetic, and polysynthetic are simplified notions of ideal languages. They are commonly used in introductory linguistic textbooks and the like. But, they do not necessarily accurately describe the situation. To be more rigorous about this you could establish a numerical value of the word-to-morpheme ratio and then make your comparisons.

3) What you are alluding to above when you say "contraction of affixes" is the degree of segmentability. Easily segmentable languages are called agglutinating, not so easily segmentable are called fusional. Young & Morgan (1987) say of Navajo verb: [it] "consists of a stem preceded by two or more prefixes, ... agglutinated together in a fixed relative sequential order...". It is easy to take issue with this. Some affixes (esp. suffixes & enclitics) are agglutinatively combined to other morphemes, but a lot of prefixes are rather fused together. It is quite complicated. Most of early Navajo linguistics was trying to figure all these "contractions" and permutations and other weird things (like changing tones).

Anyway, it's a interesting language. Maybe when Wulong has time s/he will give us some more things. Cheers! - Ish ishwar 20:50, 2005 Jan 5 (UTC)

{{ConvertIPA}}

The following has been added to this article which I removed:

This language or phonology-related article needs to be fully converted to IPA. See IPA in Unicode for information about the correct codes for IPA characters. When converting the article, please ensure that all IPA coding is surrounded by the {{IPA}} or {{IPA2}} templates. Once you've converted the article, it may be appropriate to add the {{IPA notice}} template.

I removed this for a few reasons.

1. Not everyone is a linguist. Not everyone cares about IPA.
2. Not everyone knows how to read IPA. Why learn both IPA & the Navajo writing system when Navajo writing is sufficient to describe the sound system? I used the practical orthography because it teaches how to read the orthography and it works to give a decent description of Navajo sounds.
3. If you look at most pedagogical works on specific languages they dont use IPA. Nor do general-use encyclopedias.
4. The Navajo writing system is cool because it combines English character usage with the Americanist phonetic transcription system. Many linguists favor the Americanist system over IPA, which has many annoying things about it. It is nice to give beginning wiki-phoneticians a taste of alternate systems currently in use by linguists.
5. I understand that you want to standardize. I will put a link to Southern Athabaskan languages for people who want a very linguisticy description. (actually I am torn on what to do here. how much should this be geared toward people who know a little about linguistics, how much should it be pedagogical, how hard, how much to put here vs. how much to put in Apachean languages article.....)

I think IPA should be used for the orthography and the phoneme chart, for those who are interested in linguistics and comparison (and for the fact that IPA is supposed to be just that: international. Not every Wikipedia user is American). Plus it explains the orthography: for instance, the fact that Navajo dz isn't actually voiced (AFAIK), but unaspirated, isn't clear in the Americanised system. However, beyond the orthography and the phoneme chart, I see no reason why the Navajo orthography shouldn't be used. It's standardised and current, after all, and the French article doesn't use IPA for its example sentences, right? :) thefamouseccles 02:54, 16 Jan 2005 (UTC)

 

Hi.

1. I am worried about readers who might be scared away from the weird symbols. But, if they are intrigued & sucked into linguistics because of the weird symbols, I dont want to deprive them of this opportunity. Anyway, I have used IPA in the Apachean languages article (although I am describing Western Apache mostly).

2. If this community of wikilanguagers wants to go IPA, thats cool with me. There are benefits of standardization esp. when making comparisons between langs.

3. Americanist transcription system is not used by only Americans. It is used by linguists who work on languages indigenous to the Americas. This is a legacy of the early anthroplogical linguists who created it. This system is as old as IPA, historically more stable, and also very similar to IPA. IPA wants to be international, but I think that it is not. IPA is a European creation so it already has that bias in its choice of symbols (there are sounds in Chinese languages that still dont have IPA symbols). The main issue, though, is that linguists working in specific language families often have transcription systems unique to that family. From what I can tell, this situation really hasnt changed, except in work on phonetics where IPA is consisently used in journals. And if someone wants to look into Apachean linguistics, they will have to learn 4 different systems: (1) IPA, (2) current Americanist, (3) system used from around 1935, (4) system used before 1935. I guess this is the same most everywhere. So the point: IPA is not as international as it wants to be and therefore a claim that it is international does not convince me of it being the choice for international use. (here I'm talking about linguistics, but in other language areas like dictionaries, pedagogical materials, etc. I havent seen that much standardization either. others know more about this than me, though.)

4. The Navajo orthography is different from the Americanist phonetic transcription. So, the voiceless affricate you mention would be written as [c] in this system. But, you are right about the Navajo orthography being purposely similar to English spelling practices. The symbols <b>, <d>, <g>, <dz>, etc. are used for voiceless stops because that's the way the symbols are pronounced in English. (likewise Chinese romanization was changed for English speakers' benefit much to the annoyance of non-English speakers like French who really do have voiced stops!)

Simply put, the issue is either

(a) use IPA and provide standardized unambiguous phonetic information to linguistically minded readers but complicate the presentation for more casual readers, or

(b) use only the orthography and simplify the presentation but make those interested in phonetic detail have to work harder to determine the precise phonetic values.

If phonetic precision is desired by the community then I will have to resort to phonetic transcription to accomplish this. But so far I havent seen the English or French language articles reach a very precise description of phonetics.

Your further thoughts are welcome. Thanks. - Ish ishwar 10:17, 2005 Jan 16 (UTC)

Believe me, I know the issues you're talking about. :) I work fairly intensively with the Northwest Caucasian family, in which most linguists still use a Roman-based system not too dissimilar from the Americanist system in some aspects. (And in Ubykh, my speciality, there are many phonemes that do not have specific IPA equivalents - like the Chinese languages you suggest. Even the Czech ř still has no IPA equivalent.) But I think you're oversimplifying the issue - it's not just about "IPA" XOR "Americanist". I merely suggest that in the phoneme chart, both IPA and Navajo orthography equivalents are given, and that the rest of the article can then be in the normal Navajo orthography. That way there isn't a preponderance of IPA in the article, but those who do know IPA can puzzle out how to pronounce Navajo text that doesn't have an IPA transcription. IPA isn't as international as it wants to be, but a lot of that is because so many linguists simply do not use it, choosing established orthographies instead, as you mention. (This is still a problem in the Northwest Caucasian community. It took me two years of digging and research to pin down how the phonemes of Ubykh are pronounced. If IPA had been used instead, things would have been a lot simpler.) And the reason IPA is not so stable is because it does change to incorporate sounds that are newly agreed to be phonemic or phonetic. That's the idea behind it.

And if you'll forgive my being blunt, I don't think a specimen phoneme set for a language family, even for subfamilies, is the right way to go. As an example, in the Abkhaz-Abaza subfamily of Northwest Caucasian, no dialect of Abkhaz shares more than 85% of its consonants with any other, and Abaza is different again. There's plenty of room on Wikipedia for individual phoneme sets for each language - after all, while this may be an encyclopaedia, we're looking to collect as much knowledge as possible, even for the linguists. (But I do agree with you that French and English are very poorly done; I think that this might be because linguists who regularly use IPA are less likely to concentrate on English and French, perhaps? I think many linguists of English are still using macrons to indicate long vowels... ;) (Oh, and English *does* have voiced stops; unaspirated voiceless ones are what it lacks.)

Glad to have this discussion with you though - a clash of viewpoints can often be productive. Let me know what you think.thefamouseccles 12:41, 20 Jan 2005 (UTC)

 

Nice talking to you too. I have thought about this and I guess you are right. Since most language articles attempt to include IPA transcription, it is probably best that my contributions follow suit. So, I will make the changes. I want to consult some phonetic studies by Peter Ladefoged and Joyce McDonough before doing so, however.

Some minor quibbles & thoughts:

• IPA works on a phonemic contrastivity principle. It provides a way to symbolize all phonemically contrastive sound (theoretically at least). But, it does not, generally, provide notation for every possible sound. Example: a pharyngeal trill. This sound occurs in some dialects of Haida. Others dialects have a pharyngeal fricative instead. These two sounds dont contrast phonemically, so no IPA. But they are important in a description of Haida dialects, so a non-IPA way to transcribe this must be invented. So, it is a not universal phonetic alphabet, but rather a selective phonetic alphabet — not "phonemic or phonetic", but just phonemic. There is a nice discussion of this in Daniels & Bright (1996:823-824) if you are interested.
• I retract my earlier statement about the instability of IPA. I think both European & Americanist phonetic transcription have changed a fair amount since their inceptions.
• So as far using Western Apache to exemplify the Apachean group, I think that it works pretty well. Not as much variation as in Caucasian language groups. There are historical sound changes, but these have not changed the overal phonemic inventory. One exception: Western Apache has borrowed aspirated bilabial stops from English. I dont know if Navajo has this phoneme (my work so far has been on Western Apache & Chiricahua Apache). Chiricahua has one word with this sound, but is onomatopoeic. But, I admit that I dont know about Plains Apache. I could be wrong. So the article is tentative & perpetually "in progress". But at any rate, the article before was called "Apache language" and the authors considered all "Apache" peoples and languages to be the same, so it at least it is somewhat more accurate now.
• About English stops: Most varieties of American & British English dont have voiced stops. I know that is the way they described in introductory linguistics textbooks, but this is a simplification of the phonetic reality. You will have to consult a more detailed phonetics manual to find this. The stops orthographically written as <b>, <d>, <g> are usually voiceless even within a word and surrounded by vowels (the vocal fold vibration stops soon after oral closure and starts again when the closure is released into the vowel). So, you can say that English has phonemically voiced stops that are phonetically manifested as voiceless. As usual this applies most varieties, but perhaps doesnt apply to more divergent varieties, such as Irish English, subcontinent Indian English, etc.

Again thanks for your thoughts. Cheers! - Ish ishwar 18:47, 2005 Jan 20 (UTC)

Reference:

Daniels, Peter T.; & Bright, William. (1996). The world's writing systems. New York: Oxford University Press.

Are you calling me divergent? :P Australian English (my dialect) manifests clearly voiced stops in almost every position, and that's what I was basing it on... maybe I was wrong. But it's in the nature of stops to have a very short period of flowing air. If the situation you describe is correct, then there would be no such thing as phonetically voiced stops in any language (unless they were kept very short).

Anyway, that's fairly tangential. :) I find it interesting that Western Apache has [p]... Navajo does lack that phoneme (in native vocab, at least). From what I've seen, the Athabaskan languages as a whole are poor in bilabial consonants. And I didn't think Navajo had prenasalised stops (although I am working from a 1951 grammar); I'd thought that [n] was moraic word-initially before [d], at least, and could carry contrastive tone in that position. But Navajo phonetics is fairly well treated compared to a lot of other languages, so I don't think you should have too much trouble finding material. Since Apachean linguistics is your speciality, I'll bow to your findings. :)

And I must agree; the IPA is still very defective in terms of actual phonetics. If there's a space on the IPA grid for a pharyngeal trill, and it exists in any language, why not use that empty space? But it's about as universal a system as we'll get... in the short term, at least. thefamouseccles 00:47, 21 Jan 2005 (UTC)

I once heard an Australian speak her native English for hours. It was a mind-boggling experience. As described above, her intervocalic /b d g/ were devoiced in the middle: the voicing stopped after the stop began, and resumed before the release. I find myself unable to imitate that.

I guess the IPA will get a symbol for an epiglottal trill (which is what the Haida phenomenon is – not pharyngeal) in its next revision. All over Wikipedia it's mentioned that the symbol я is occasionally used in the literature.

What is really annoying about the IPA is the tradition of writing one-contact alveolar trills the same way as flaps – but that's not strictly speaking a failure of the IPA itself.

David Marjanović | david.marjanovic_at_gmx.at | 22:26 CEST | 2006/5/7

 

(I guess we are getting a little off topic Navajo-wise. Oh well.)

Yes just about all Athabaskan have few labials. In fact Tlingit does not have any labials (stops, fricatives, etc.) — it only has a labio-velar glide & labialization as a secondary place of articulation. So, Tlingit has disproved the claim that labials are universal. Nice! It is interesting to note that a Salishan language, Tillamook (spoken in Oregon, USA), does not have labials either and it is surrounded by Athabaskan languages (other Salishan languages do have labials). Additionally, there are probably not many words that contain labials, esp. [m]. For instance, in my data on Chiricahua Apache there are 3 stems in native words that have the [m] phoneme and 3 other words borrowed from Spanish. Not many words with [m] (more words have [b]). But, to make up for the lack of labials, Athabaskan languages are coronal-heavy.

One other Athabaskan language, Hupa, surprisingly lacks velar stops!

I wouldnt guess that your variety is very divergent. Usually, the consonants are fairly similar in most Englishes. You are right that keeping voiced stops short in duration would be physically easier, and because of their greater physiological effort they are less common than voiceless stops cross-linguistically. But I do believe that there are some languages with geminate voiced stops (I dont know which). Stops can be voiced in several ways, two examples are (1) allowing expansion of the cheeks & throat, or (2) lowering the larynx to accommodate the increasing air pressure during vocal fold vibration. As I mentioned above, French is a language with true fully voiced stops. It is quite common that stops that are phonemically voiced (i.e. they pattern like the other voiced consonants) are realized as phonetically voiceless. I'll look up some references & put them on your talk page.

Different languages & dialects have different things for the /nd/: [nd], [d], [n], [n^d], and [ⁿd] are all possibilities. Simplified, Navajo has [n], Chiricahua has [n^d], Mescalero has [nd], Western Apache has [nd], [d], [n], [n^d], & [ⁿd], Jicarilla & Lipan have [ⁿd], Plains Apache has [d]. Additionally, some languages/dialects have a similar situation with /m/ I going to put some things about this in the Apachean article one day.

Peace. - 07:26, 2005 Jan 22 (UTC)

łéé'íí'nííł 'cemetery' ?

I dont find łéé'íí'nííł in Young & Morgan's (1987) dictionary or Young et al.'s (1992) lexicon. I do find for both łeeh hwii'níłí 'cemetery, graveyard (lit. 'the place where they are buried one after another')' [łeeh 'into ground' < leezh 'dirt, ground, soil' + -iih 'into'; hwii'nííł 'they are put one after another' < -nííł 'to handle PLO1'; -í (relative enclitic)] and łeeh ho'dii'nííł 'graveyard, cemetery, burial ground'. (There is also jishcháá' = 'grave, graveyard, cemetery'.) However, there is a form łee' 'inside/within the ground' [ < leezh + -ii' 'inside']. Can the person who provided this form please state the source? I am interested.

Since I dont find this form, I am removing it and replacing it with other words which I hope are equally interesting.

Thanks. - Ish ishwar 18:31, 2005 Jan 12 (UTC)

I provided this form, which can be found in Gladys Reichard's Navaho Grammar (1951; Publications of the American Ethnological Society) as an example form for the prefix łéé' in the ground. This may be an idiolectic usage, and I have no issue with you replacing it, but it is (or, perhaps, was) real Navajo. thefamouseccles 02:34, Jan 16 2005 (UTC)

Cool. Thanks for the note. I havent gotten my hands on this one yet. I have been told her grammar is good but sadly neglected. Also heard that there are many words missing from Young & Morgan's work — not hard to get a word from a native speaker that isnt there.

Young & Morgan also list a prefix łe'- that means the same. Peace - Ish ishwar 08:48, 2005 Jan 16 (UTC)

No worries. I've come across this in Wichita too, where foreign nouns are often construed with verbs - very similar to Navajo, and just as hard to write a definitive dictionary for. ;) Reichard's grammar is good, but being from 1951 probably not the best for modern Navajo usage. The transcription is good though - I recommend it if you can find it.thefamouseccles 13:22, 20 Jan 2005 (UTC)

Grammar: Nouns

OK, two issues here, so I'm going ahead and dividing this again:

POV or General Comprehensibility?

The following was added by Barfooz at 20:40 on 24 April 2005:

Many concepts that exist as nouns in other languages exist as verbs in Navajo. Noun phrases exist in Navajo outside of syntactic space: that is, they are not necessary for forming a grammatical sentence and exist purely for semantical reasons. Noun phrases are unique in Navajo because they lie in the adjunction domain, rather than in A or A’ positions, and do not receive overt case marking.

As I pointed out nearly a year ago, this explanation unnecessarily relies on knowledge of Chomskian syntactic theory to be comprehensible (if even then). I propose rewriting it thus:

Many concepts expressed using nouns in other languages appear as verbs in Navajo. The majority of true nouns are not inflected for number, and there is no case marking.

Noun phrases are often not needed to form grammatical sentences due to the informational content of the verb.

I believe this is clearer and will be more easily understood by a general reader. If there are no dissenting opinions, I will change the article within a few days. RJCraig 04:58, 26 March 2006 (UTC)

There being no dissenting opinions in the ensuing weeks, I have made the change proposed above. RJCraig 15:12, 9 April 2006 (UTC)

Nouns section in general

Surely there is more to say about nouns than this? Discussion of the personal possessive prefixes, for example.

RJCraig 12:09, 21 March 2006 (UTC)

yes, there is more. I thought that I wouldnt update this article until I update the general article, where you can find a start of a discussion of nouns: Southern Athabascan languages#Nouns. There is much missing here such as the stem alternations of possessed nouns, and the difference between inherently possessed and non-inherently possessed noun stems.

Thanks for the feedback; some mention of the limited inflection for number on nouns dealing with humans might be in order as well; I'm thinking of the -ké suffix here. (BTW, the similarities between Navajo and Japanese nouns are very interesting.) Any thoughts as to my change proposed above? RJCraig 04:58, 26 March 2006 (UTC)

??help

i am trying to figure out the correct pronunciation for "mosi/mósí", navajo name meaning cat. anyone able to help me??

Hi. mósí is the word for cat. mosi probably isnt a Navajo word. The consonants m and s are pronounced the same as in English (which i assume is your native language). ó is pronounced similar to the vowel in go or mode but with a high tone. í is pronounced like the vowel in see or tea but with a high tone. So both vowels are pronounced with high tones — the accent mark ( ´ ) represents a high tone. If you dont know, Navajo is a tonal language (like Chinese, Thai, or many Bantu languages). High tones are pronounced with a higher vocal pitch than low tones. There is another similar word that means cat which is másí and an unrelated word that also means cat: gídí (which is borrowed from English kitty). Also you can specifiy a male cat by adding something to the word, giving you mósíką’. happy languaging — ishwar  (SPEAK) 05:41, 2005 Jun 6 (UTC)

howdy! i agree with ishwar with the above pronunciation and definitions. Mósí also has a lax variant as well, mʊsɪ (with high tones)— User:Daghaalsuii

Fortis and lenis

In the consonants table b, d, g, dz and j are explained as [p], [t], [k] and so on. This is clearly good enough for a phonemic transcription. But the audio samples from Ladefoged (link at the bottom of the article) show that these sounds are voiceless lenes: [b̥], [d̥], [g̊] etc.. Voiceless lenes occur elsewhere, such as in Mandarin (Pinyin b, d, g) and at least some kinds of Spanish (p, t, c/qu). Unlike these examples, there is a phonemic difference between voiceless lenes (b, d, g) and unaspirated fortes (p, t, k) in most southeastern kinds of German, including the Austrian version of Standard German (and my dialect).

Which nits should I pick? Should I replace [] with // in the consonants table? Or should I put the correct voiceless lenes into the brackets that are there?

While I am at it, should I replace "unaspirated" with "plain"?

Oh, and is there a rule when h is pronounced [h] and when it's [x]? The Ladefoged example has a short [x] like that of Mandarin, but it's just one example...

David Marjanović | david.marjanovic_at_gmx.at | 01:50 CET | 2006/2/14

hi. What's your definition of fortis/lenis? I find this terminology to be more of a phonological opposition if you are speaking in terms of different VOTs.

h/x is [x] when stem initial and [h] when prefixal or stem/word final. Other things happen at the stem boundary, like increased closure duration of stops. Related Western Apache /t/ (orthographic d) flaps between vowels except when stem initial.

– ishwar  (speak) 00:09, 15 February 2006 (UTC)

Thanks for the prompt feedback!

Definition?

1. I know it when I hear it. (Yeah, OK, they are extremes in a continuum... but they seem to be more common than the middle. Perhaps Romance /k/ is in the middle.)

2. Intensity, air pressure, is different, as the names suggest.

3. The easiest and therefore most common method for increasing air pressure is to prolong the hold of a stop. Indeed fortes are almost always longer than voiceless lenes.

Examples: All plain stops and affricates here are voiceless lenes. I would say that the word-initial plain stops here are voiceless lenes while the word-internal ones are fortes, but I'm not quite sure about all of them (no surprise because that's not a phonemic contrast in Lakhota). The over-aspirated stops sound like voiceless lenes followed by [x]. Quechua has only fortes, even though it has the same plain/aspirated/ejective phonemic contrasts as Navajo and Lakhota.

Bullshit! Was I even more tired than I am now, or what? The word-initial /k/ which occurs in one of these examples is a fortis. All other plain stops there are lenes, including the word-initial /q/. I guess I had forgotten all except that one initial /k/. grmpf The one /tʃ/ (word-initial) is also a fortis.

David Marjanović | david.marjanovic_at_gmx.at | 23:39 CEST | 2006/4/3

Thai, on the other hand, has an interesting phonemic contrast between voiced and voiceless lenes (and aspirated fortes), as has Spanish (according to my very limited experience with it) but not French.

Voice onset time is identical in voiceless lenes and in plain fortes: both are tenues.

Thanks for the explanation of h. However, in the audio samples word-final /h/ is [ɦ̥], assuming I'm really hearing and not just imagining it.

"Increased closure duration of stops" should usually produce fortes, so I gather Navajo has plain fortes at stem boundaries. But of course all spoken Navajo I've ever heard are the 13 words at the link above!

David Marjanović | david.marjanovic_at_gmx.at | 00:10 CET | 2006/2/18

hi again. long time, no answer. But here is a short answer.

Basically, the fortis/lenis contrast is a phonological contrast. These terms mean different things phonetically depending upon the language you are discussing. They were originally used in by German dialectologists and are becoming somewhat outdated within much phonetic literature, but are still commonly encountered in general stuff.

The contrast is variously between

1. voiced - voiceless
2. unaspirated - aspirated
3. less expiratory/articulatory energy - more expiratory/articulatory energy (i.e. intra-oral pressure and/or muscular tension)
4. shorter closure duration - longer closure duration (in stops)
5. shorter first formant (F1) transition time - longer F1 transition

I am talking about the third. However, southeastern German is the only, mmm, part of a language I know of where this alone is phonemic, so everyone tends to overlook it.

Found another: Click on the examples for j and ch in Haida. Both are voiceless and unaspirated. After some initial surprise I've found out I can distinguish them – j is a lenis and ch a fortis.

David Marjanović | david.marjanovic_at_gmx.at | 21:50 CET-summertime | 2006/5/7

All other languages that have fortes and lenes in the first place seem to use additional features (voicing and/or aspiration). Not all combinations are possible – for example I can't manage to pronounce a voiced fortis stop.

Spanish (or at least the kinds I've heard), by this definition, lacks fortes – it has voiced and voiceless lenes, thus voicing alone is phonemic there. Quechua (in the audio sample mentioned above, anyway) lacks lenes – it has plain, aspirated and ejective fortes. (Aspirated voiceless lenes are most probably impossible; ejective lenes seem to occur in the Caucasus – there's a slightly weird audio sample of Abkhaz out there; I'll try to find it – but not in America.)

Obviously this is a difference of degree rather than kind. In languages where neither this difference alone nor another (such as voicing) that requires lenes or fortes is phonemic, we can thus expect free variation throughout the spectrum. This explains why some of the Lakhota samples confuse me and why in these Tlingit samples I find that one speaker makes unmistakable fortes most of the time while the other prefers lenes for the same phonemes.

The simplest way to generate a fortis stop (not a fricative, of course) according to this definition is to prolong the hold, so that more pressure builds up. Indeed southeastern German fortes are almost always longer than lenes. However, long lenes are possible, and IMHO occur in some of the Navajo samples.

And in Lakhota. Well, I can't tell if the "long" or the "half-long" IPA symbols would be more appropriate, but… :o)

David Marjanović | david.marjanovic_at_gmx.at | 21:50 CEST | 2006/5/7

I can't comment the 5^th because I don't know what the formants are.

Correction: Formant says formants only occur in sonorants, so I guess it can't matter here.

David Marjanović | david.marjanovic_at_gmx.at | 14:30 CET-summertime | 2006/4/1

The voiced-voiceless and aspirated/unaspirated contrasts can be described acoustically as having different average measurements of voice onset time (VOT), which is the time of the periodic vocal fold vibration with respect to the release of a stop. Closure duration is the amount of time that the mouth is closed during the articulation of stops. The F1 transition time is the measurement of time that it takes for the first vowel formant to reach a steady frequency. The muscular tension/articulatory energy contrast is also called a tense/lax contrast and is practically impossible to measure consistently & therefore not very useful.

Voice onset time doesn't seem to be a very precise criterion for aspiration. My /k/, as that of Quechua, is a bit exaggerated, resulting in slightly delayed VOT, but apparently not in any aspiration whatsoever. On the other hand, strongly aspirated consonants can be pronounced so fast that, for what that's worth, I can't hear a VOT delay. (Sample: the first Harry Potter film. "Scared, Potter?" – "Yyyyyou wwwish.")

The tense-lax contrast in consonants seems to be nothing but length, but the Wikipedia page on this is my only information...

Because the exact meaning of terms such as fortis and lenis are vague and have different meanings in different contexts, they are dispreferred in phonetic description. However, they are often useful in phonological description since you may want to have a cover term that will cover things that are different phonetically but similar phonologically.

So, basically I find these terms objectional in a precise description of Navajo.

I have nothing else for how to call the third of the differences you mention above. On the other hand, this difference is quite obviously not phonemic in Navajo, so entirely superfluous in a phonological description of that language.

I still havent listened to the sounds files that you mention, but I'll do so some time in the future.

Please do.

– ishwar  (speak) 18:19, 22 March 2006 (UTC)

David Marjanović | david.marjanovic_at_gmx.at | 23:28 CET | 2006/2/29