Talk:Indo-European ablaut/Archive 2

POV/Peacock/Confusion

The article is full of strong or emotive terms and un-attributed claims, with no in-line references to support. Despite having a list of references at the end, it reads like OR. Examples below.

• its far-reaching consequences
• must be distinguished
• is crucial also to notice
• atypically neat example
• may never be found anyway
• It has been shown that it seems to be highly likely
• it is not difficult to imagine this
• has been said to be due to
• few reflect the original system as neatly as Greek
• is obviously related to Latin
• which we know
• some scholars reconstruct
• they must have been significant
• what most people primarily associate

-unsigned comment by Orange Dog

This is a very complex topic, and you yourself said you find it confusing. Yet this is in fact the most user-friendly explanation for beginners that is currently available anywhere. We can rightly be proud of it. Now you are criticising the very things which make it easily understandable as "peacock language". This is crazy.

First of all, POV is a meaningless claim unless there is controversy. Although scholarship is still working on some of the detail, this is not a topic where there are two opposing points of view. Therefore the article cannot be taking sides. If you are going to accuse someone of POV you have to explain what perspective they are biasing, and I doubt you can do that here.

The particular phrases you have picked out seem fine to me. If we give a example from Greek, which is chosen because it shows the pattern neatly, what possible objection can you have to us adding the warning that this is untypical, and that most examples in most languages are less neat? If a phenomenon has consequences for the struture of over a hundred languages, what can possibly be wrong with calling these consequences far-reaching? These claims are objective and fair. And some of the other phrases you cite are just the function words which make an English sentence work. Some of these phrases could be improved stylisitcally - feel free to do that - but none of them are peacock language.

What several of these phrases do represent are claims which need to be verified. So your other complaint about in-line citations is fair enough. But I would suggest you leave it at that. Having four caveat boxes at the top gives the impression that the article is highly problematic. It is actually a very strong article, well written and well informed, which needs only very minor improvements. --Doric Loon (talk) 22:23, 16 January 2009 (UTC)

I concede that this article presents a very good explanation, but I contest that it's tone is suitable for an encyclopaedia. Clearly there are some issues, exemplified by failure of a GA review. When I first read it I found it very much like a lecture and was going to tag for WP:COI before I checked the referenced authors and edit history. I have attempted to improve this (re-wording 1st person "we can see", "we know", etc.) but didn't want to go any further as I am not an expert on the subject, nor do I have ready access to the references. The phrases I have highlighted above all either unnecessarily employ emotive language, or make claims without citation. For example, either the example is "related to Latin" (in which case please provide a verifiable source), or it isn't. Whether this is "obvious" or not is highly subjective. Similarly, whether noticing something is "crucial" or not is unlikely to be verifiable, or indeed true for most people. This article could be greatly improved by removing/rewording all such language and adding in-line citations where appropriate. Given that (by the article's own admission) the subject is still a topic of research/debate, it may be best to omit some controversial claims, or qualify them as not having a consensus. –OrangeDog (talk • edits) 15:45, 17 January 2009 (UTC)

Addendum: I've just noticed User:Angr has already re-written[1] some of the examples I have highlighted, but there is still room for improvement, especially in supplying citation. –OrangeDog (talk • edits) 15:51, 17 January 2009 (UTC)

Yes, thanks Angr, that did help. You're right, OrangeDog, that this is stylistically nicer. Please do make more such changes if you think it will help. It is not surprising that you think this sounds a bit like a lecturer - Angr and I are both university teachers. I never heard before that that was a bad thing, but i am not attached to the tone. --Doric Loon (talk) 18:35, 17 January 2009 (UTC)

I should also have said a word about your other complaint - confusing. I think you need to consider carefully whether you are confused because it is a difficult topic, or because it is explained confusingly. I am confused when I read wiki articles on pure maths, but if they carry GA tags, I have to assume that the problem lies with me - because no matter how well the articles are presented, you will have to work to understand a complex topic in a subject you have never studied.

However, if you think our explanations could be less confusing, you are welcome to give a detailed critique. What is not acceptable is just to throw a "confusing" tag at the top of the article and not say what you find confusing. --Doric Loon (talk) 22:31, 16 January 2009 (UTC)

The main point of confusion I think is the use of PIE in examples, without any explanation of notation/pronunciation. This renders the examples opaque as they're being used to illustrate pronunciation. Specifically, what do PIE vowels sound like and why does every work start with '*'? If there is no verifiable consensus on PIE pronunciation, these examples should probably be removed.

A second suggestion: having a separate section explaining the grades may be helpful instead of the current arrangement with zero- and a- grades having their own section and e- and o- incorporated into general discussion. These sections need not be long, possibly as subsections of PIE Grade system. –OrangeDog (talk • edits) 15:45, 17 January 2009 (UTC)

The problem of not knowing how to pronounce PIE is a real off-putter for beginners. When I was a student it bothered me enormously that PIE forms were not pronouncable words I could learn and use. But that's the way it is, and we have to work with it. There is no way we can get around that, because the PIE roots are not just "examples", they are the foundation of the whole argument. If a Latin example doesn't work for the reader, we can take one from Greek, but we can't discuss any of it seriously without the PIE.

I don't have an answer to the question of how much basic knowledge of PIE we have to put into every article which mentions the topic. There are after all dozens of them. Do we have to say in every one what the asterisk means, or can we expect that the interested reader will follow the link to the PIE main article to find out. Ideally one wants to explain everything right here on the page, but if all basic knowledge has to be explained on every page, we will never get to the more advanced stuff. Going back to the maths example, wiki articles on advanced maths concepts don't explain all the basic maths which underlie them - they just point to other more basic articles. I think the trick is to be as helpful as possible but not to the extent of cluttering the article. --Doric Loon (talk) 18:35, 17 January 2009 (UTC)

If someone could find a vague mapping from PIE vowels to IPA in a source somewhere and add it I think it would be very helpful, again given that this topic is about vowel sounds. (The PIE article itself doesn't say what the asterisks are about.) The main problem remains lack of in-line citations though. –OrangeDog (talk • edits) 19:19, 17 January 2009 (UTC)

The asterisks indicate that a form is reconstructed, i.e. it is not recorded. Since PIE was never written down, all PIE forms get asterisks. It's just a way of reminding ourselves that it is all a little hypothetical.

I'm afraid you are unlikely to find any serious linguist mapping PIE to IPA: the point is that we truely have no way of knowing how these sounds were pronounced. If Latin and Greek and Sanskrit all have a /p/ and Germanic has an /f/ and Celtic has Ø in the same position in lots of words, the comparative method allows us to deduce that there is a common ancestor in PIE. We can never truely know what that ancestor was, but we can guess it might be /p/ because that is what has survived in the widest range of daughter languages, and because the ones which have changed would be explicable and in some cases even produce a very neat pattern of shifts if we view them that way. So we call it *p. But actually, when we talk about PIE *p we are using the asterisked letter *p as a placemarker for a phoneme the existence of which has been demonstrated, but the actual value of which will always be uncertain. In fact, some people would say that PIE reconstructions are not so much an attempt to uncover the parent language as a way of systematically encoding the relationships between the daughter languages.

So if you do find a source which gives you a PIE pronunciation guide in more than the most tentative terms, it will be rejected as unscholarly. Sorry. It would be lovely to be able to make this more like a language, but it is not, it is a very abstract coding system, and we have to live with that. --Doric Loon (talk) 21:23, 17 January 2009 (UTC)

Hmm. It would seem that this article needs a brief summary of these points at the beginning. As far as I see it, there are a couple of issues that need justifying and are sources of confusion.

• Phrases such as "from which the English word [is] derived" become ambiguous/wrong as technically PIE is derived from its daughter languages.
• How can one talk about pronunciation differences of vowels using PIE if any attempt to pronounce PIE is seen as "unscholarly" -N.B. pronunciation of consonants is largely irrelevant to this article
• Why is PIE being used at all, when equally applicable and more accessible examples can be found in "modern" Indo-European languages (the title of the article)

–OrangeDog (talk • edits) 21:41, 17 January 2009 (UTC)

Is this the kind of explanation you are looking for? Linguists infer from the daughter languages that the daughter languages are derived from a proto language that had the vowel gradation patterns described in the article. Although we don’t know exactly how the vowels were pronounced in the proto language, the vowel gradation of the proto language is reflected in the daughter languages in predictable ways (as indicated by the spelling in the daughter languages). In particular in most daughter languages the e-grade is reflected in a front vowel (often a mid front vowel), and the o-grade is reflected in a back vowel (often a rounded mid back vowel).—Preceding unsigned comment added by TEB728 (talk • contribs) 23:12, 17 January 2009

That what I was looking for. Put that in the article, add some in-line cites, maybe re-arrange some sections and it shouldn't fail GA again. –OrangeDog (talk • edits) 23:56, 17 January 2009 (UTC)

The reason we have to discuss PIE here is because Ablaut was a regular system in PIE, which survives as irregular remnants in all the daughter languages. Only PIE shows the regular pattern which explains all the irregularities. The point is to show why all these languages do funny things with vowels.

I have no objection to any of those explanations going into the article, but I do wonder where they would fit in without detracting from the clarity of the main material. Perhaps at the end of "preliminary considerations"? --Doric Loon (talk) 16:26, 18 January 2009 (UTC)

OK, I've added a short paragraph summing up what has been said here. Please play around with it to make it as easily understandable as possible. But please do not let it grow longer and longer with all possible details. We do want it to be understood, but mostly this article should build on the basic knowledge which is explained in other articles. --Doric Loon (talk) 16:54, 18 January 2009 (UTC)

Thanks for taking the time to contribute to this discussion. I've had a quick look at other PIE-related pages and would like to point out that it should always be made as clear as possible whether one is talking about Indo-European languages in general, the unknown common ancestor of these languages, or the theoretical reconstruction of that language (vis. PIE). –OrangeDog (talk • edits) 19:41, 18 January 2009 (UTC)

Reorganizing article

I am thinking of reorganizing the article somewhat.

• Moving the etymology paragraph from the lead to the “Preliminary considerations” section The reason for this change is that it is not a summary of anything; so it doesn’t belong in the lead.
• Re-titling the “Preliminary considerations” section and moving it to bottom of the text The reason for this change is that Apophony has taken over role as the main article on vowel gradation.
• Making “Zero grade,” “a-grade” and “Ablaut and grammatical function” subsections of “Ablaut in Proto-Indo-European”
• As a result of this subordination, moving “Subsequent development of ablaut” below “Ablaut and grammatical function” —teb728 t c 03:59, 11 May 2009 (UTC)

I always feel that the place to say where a term comes from, if it is a brief statement rather than a discussion, is right up at the first mention. It is part of the basic definition. So I would leave that "etymology" where it is. Otherwise it will get lost in sections further down which are more theoretical. And I would resist your second suggestion. The Preliminary Considerations section was added in response to requests from people on this talk page who found the article difficult to understand and said it was unsatisfactory that they had to go to other pages to get the tools they needed to understand what we were talking about. We have tried hard to satisfy those needs and the result seems to be a text which beginners can make sense of. So I think that starting with an explanation of basic ideas is good. Calling it Preliminary Considerations means that people who don't need it will recognise that they can skip it. Besides, the Apophony page cannot be trusted to provide the preliminary information required since it is unstable, and if you go back in its history you will find that it takes a synchronic approach which linguists will find complementary to our diachronic approach, but which will really confuse beginners if they are sent there for an explanation of what they don't understand here. I have no objection to your third suggestion. I don't see the point of the fourth suggestion. Surely it is better to deal properly with the purly phonological questions first, before getting into the more complicated question of their implications for morphology? I am a little afraid that if you put the morphological section first you will end up giving the misleading impression that Ablaut in German verbs is the logical starting point for understanding this. --Doric Loon (talk) 20:14, 11 May 2009 (UTC)

Still incomprehensible example

Nice article, folks! But there was one thing I was unable to follow: the reasoning behind the "*deywó- / NPl. *-es" example, which is mentioned twice. I see that this was already brought up last year, and identified as a problem, apparently easily solved by anyone with access to the sources cited in the article. At the very least, could someone add a gloss for *deywó and expand the abbreviation "NPl."? I also found the final sentence of #Zero grade dismayingly impenetrable:

ablaut is seen as originally consisting of a combination of vowels which form a transfixal and discontinuous vowel melody

The scare italics and redlink are a particularly nice touch… CapnPrep (talk) 03:42, 30 November 2009 (UTC)

Rewrote the paragraph, though I don't have access to the original source. I may have misunderstood some things, but at least the wording is less than hopelessly obscure. — Erutuon (talk) 16:20, 7 December 2009 (UTC)

What is all this on Semitic root+transfix morphology? Is it just sb's personal imagination (I can't see any references or notes) and is there any evidence for that? AFAIK PIE morphology is strictly suffixal (there is only one infix for 1 type of presents). --Ivan Štambuk (talk) 17:47, 7 December 2009 (UTC)

Lengthened grade

The example *ph2-tēr, nominative of 'father' in the lengthened grade, isn't really a distinct grade at all in this case. Rather, it goes back to an older *ph2-ter-s, which lost the -s and lengthened the vowel through operation of Szemerényi's law. Stang's law had a similar lengthening effect as well. So this isn't really a good example because it doesn't show a distinct grade, but rather the result of an older sound change. CodeCat (talk) 11:37, 29 July 2011 (UTC)

I tried to point this out but got reverted; unfortunately I never found a really convincing example. Sihler (who has some disputable views) says somewhere that there were no lengthend grades at all in PIE, only lengthening effects through laryngeals, Szemerényi's, etc. Can you come up with a good example for a lengthened grade? --ἀνυπόδητος (talk) 11:47, 29 July 2011 (UTC)

Germanic has a lengthened grade in the past plural of strong verbs of class 4, 5 and the entire past of class 6, and maybe the past of class 7 as well. But I don't know if those have convincing PIE origins. CodeCat (talk) 14:02, 29 July 2011 (UTC)

AFAIK Balto-Slavic is the only branch that distinguishes lengthened grade vowels and long vowels of laryngeal origin, specifically in acute/circumflex accentual distinction, which is still alive & kicking in modern-day Serbo-Croatian (short/long) and Lithuanian (falling/rising) on the inherited vowels.

Are you sure that the specific origin of the long vowel matters at all when classifying it as lengthened-grade? If it was indeed a genuine PIE */ē/, why does it matter whether it came about by a specific compensatory lengthening in some pre-PIE period, or as a regular morphological sound change?

The problem is not in listing a specific example for lengthened-grade, but in listing a single root demonstrating all the ablaut variations. --Ivan Štambuk (talk) 11:03, 31 July 2011 (UTC)

It doesn't matter as such, but while e, o and zero grade have very specific grammatical origins, the lengthened grades are much more haphazard, so it's harder to explain how they fit into the ablaut system. And it doesn't help either that the lengthened e and o grades are in fact the outcome of regular short e and o grades that have undergone lengthening through sound change. So when speaking of the 'origin' of ablaut in the earliest PIE, the lengthened grades don't really make as much sense. CodeCat (talk) 17:25, 31 July 2011 (UTC)

The pater example is a very useful one which is often used in teaching, so I would be reluctant to see it deleted, but you can certainly add a footnote making this point, and if you can find a better example to accompany it, that would also be great. --Doric Loon (talk) 21:12, 31 July 2011 (UTC)

I just realised that there are acrostatic nouns showing lengthened e-grade in the direct cases and e-grade in the oblique cases. An example is wikt:Appendix:Proto-Indo-European/mḗh₁n̥s. And then there are the so-called Narten presents, which show the same kind of ablaut but in verbs. I can't think of any root that has all of those grades, but I still think taking examples from different roots is preferable to showing an example that isn't an ablaut grade at all. CodeCat (talk) 10:51, 10 September 2012 (UTC)

Ablaut in Nouns

It is quite obvious that PIE ablaut had a role in vebal conjugation, but what about nominal declension? The general idea seems to be that one grade was present in the Nominative, Vocative and Accusative and another in the other cases, but how many IE languages have actually preserved some form of this system? — Preceding unsigned comment added by 101.166.150.53 (talk) 23:29, 17 December 2012 (UTC)

There may be a few languages that preserve a few relics. Proto-Germanic is thought to have preserved ablaut in nominative *tanþs ~ genitive *tundiz; the former gave rise to the forms in the North and West Germanic languages, the latter is found in Gothic. There are also two nouns, fire and water, that preserve the r/n suffix alternation, reconstructed as nominative *fōr, *watōr ~ genitive *funiz/fuiniz/funiniz, *watiniz. West Germanic has -r in both, North Germanic has -r in "fire" and -n in "water" and Gothic only has -n. CodeCat (talk) 01:42, 18 December 2012 (UTC)

Ablaut and transitivity

I'm curious to see that this article makes no mention of how ablaut may generate verb pairs with different transitivity, such as English sit, set and lie, lay. Or is this some other phenomenon and not actually ablaut? -- Eiríkr Útlendi │ Tala við mig 17:58, 27 May 2014 (UTC)

That's Germanic umlaut. — Eru·tuon 18:22, 27 May 2014 (UTC)

Thank you! -- Eiríkr Útlendi │ Tala við mig 07:29, 28 May 2014 (UTC)

Preliminary considerations

Alarichall I saw you had added this material in this edit: [2].

In the section Indo-European ablaut#Preliminary considerations, in the first paragraph we read:

• (quantitative gradation: photograph and photography).

In the third paragraph of the same section we read:

• photograph/photography)

this last one in Edit Mode: ''phot'''o'''graph''<!-- the second o of photograph is bolded because it's the one that changes pronunciation -->/''phot'''o'''graphy'').

In the first instance of "photograph" (first paragraph), the first "o" is in bold, and in the second instance of "photograph" (third paragraph), the second (middle) "o" is in bold. I just wondered if these were supposed to be different, and, if so, why. CorinneSD (talk) 16:26, 7 July 2015 (UTC)

Note: @Anypodetos: It was you who (re-)added the HTML comment here, apparently missing that both Os change pronunciation. --Florian Blaschke (talk) 16:54, 7 July 2015 (UTC)

You're right, I completely missed that. So I suppose both o's should be bolded? --ἀνυπόδητος (talk) 17:55, 7 July 2015 (UTC)

I don't think so; it would confuse the reader only more. I know I don't quite get the part, and I'm a trained Indo-Europeanist! Perhaps a better example can be found. --Florian Blaschke (talk) 18:28, 7 July 2015 (UTC)

It is true that both 'o's change pronunciation, but an important rule of this kind of article is not to confuse or clutter things by showing every possible example. The first 'o' is the chosen example, which is why it is bolded. (The second one could have served just as well - but I see no advantage in that.) When 'photograph' becomes 'photography', the shifting stress pattern of English causes this long (diphthongized) 'o' to be shortened. That's all the example is meant to show. As a lead-in to PIE Ablaut, the preliminary considerations want to show that vowels can change length, change quality or reduce to zero. Yes, I know, the 'o' in photograph changes quality as well as quantity, but the point is not to give a complete description of English - it is simply to give one English example of each of the changes which will then be seen in PIE, and this was our example of quantitative changing. If you can think of a better example of quantative reduction in English, go right ahead. --Doric Loon (talk) 12:45, 8 July 2015 (UTC)

What was confusing was the way the section used at first the change of the o after ph as an example, and then of the o after t. Now that you've restored the original example, it is clear to me! --Florian Blaschke (talk) 16:39, 8 July 2015 (UTC)

Yes, I think somebody who didn't get the point had been tinkering with it. It was right when we first drafted that section.--Doric Loon (talk) 20:25, 9 July 2015 (UTC)

Thanks for your work here, guys. I think I merely added the 'photograph/photography' example as a replacement for an earlier example 'man/woman' (which I thought was even less clear!), which no-one seems to want to revert, so no complaints from me! And the current article looks okay. Having said that, the current text 'Some involve a variation in vowel length (quantitative gradation: photograph and photography)' is a poor example, because it is not purely an example of quantitative gradation. If anyone can think of a standard English example of purely quantitative gradation, that would be excellent! The closest thing I can think of is when a voiced fricative causes lengthening of the preceding vowel (so 'who' has a shorter vowel than 'whose', at least in my southern British English), but I don't know if that is a relevant/general/clear enough example. Alarichall (talk) 13:54, 13 July 2015 (UTC)

On further reflection, I've added in IPA transcriptions, including stress markers (using the Oxford English Dictionary's British English transcriptions), to make the examples of gradation caused by modern English word-stress patterns clearer. I hope people think this is useful. So instead of 'the results of English word-stress patterns (man/woman, photograph/photography)', we now have 'the results of modern English word-stress patterns (man ['man]/woman [ˈwʊmən], photograph [ˈfəʊtəɡrɑːf]/photography [fəˈtɒɡrəfi]).' Alarichall (talk) 14:10, 13 July 2015 (UTC)

Alarichall I have only an amateur interest in linguistics (just a few graduate courses), and I'd really like to understand this subject better. In a quantitative gradation, wouldn't the best example show only lengthening (or shortening) of the vowel and not an actual change in the vowel sound? In the changes in the examples above, there is not only a slight lengthening but also a change in the vowel sound. For example, in the change from "man" to "woman", the "a" sound not only becomes shorter but becomes a schwa, which is actually a different vowel sound. In the change from "photograph" to "photography", while the first "o" in "photograph" is a real "o" sound, the first "o" in "photography" is not really an "o" sound anymore; I would think it would be a schwa. (I remember learning that all unstressed vowels in English are pronounced as a schwa – ə. If the change from a stressed vowel to the same vowel unstressed is an example of quantitative gradation, then there are many, many examples in English.) I think there are very few words in English where a vowel sound lengthens (or becomes shorter) in length but keeps the same, or almost the same, sound. I thought of a pair of words where this might be illustrated: steer and steerage, where the "ee" sound becomes shorter but is still the same sound. You may think of other examples like this. Please tell me if I have misunderstood "quantitative gradation". CorinneSD (talk) 23:27, 13 July 2015 (UTC)

Hi. No, I don't think you've misunderstood the concept. The trouble is first that in English (as with every language I've ever learned - this may be a near universal) long-short vowel pairs always have a qualitative as well as a quantitative difference, second that in English (unlike say German) the awareness of vowel length is not high, and third that in standard English (unlike my beautiful Scottish accent) all long vowels are diphthongized. So it's hard to get examples. And you certainly don't want to clutter this section with explanations about the nature of English. But most people would understand that /əʊ/ and /ɔ/ are a long and a short version of <o>, because the spelling makes us sense that they are related. That is better than reduction to schwa, which is a more extreme reduction of any vowel. How about phonics and phone as an example? --Doric Loon (talk) 07:26, 14 July 2015 (UTC)

a-grade

• "... which might help to explain the vowels in class 6 Germanic verbs, for example."

I'm not really sure what needs to be explained? Class 6 strong verb vowels comes from zero-grade -(e)H₂/₃- or -o-. Anglom (talk) 20:24, 24 August 2016 (UTC)

But why the zero grade, when all other classes have full grade in the present? The verbs *draganą, *flahaną, *grabaną, *hlaþaną, *lahaną, *malaną, *slahaną, *wadaną, *wakaną and *þwahaną can't be explained as direct descendants from a zero grade with a laryngeal, because you'd get a syllabic sonorant that would surface with epenthetic u- (e.g. a hypothetical *ghrHbh- > *ghurbh- > *gurbaną). O-grade in the present tense seems even less likely as it goes completely against what is known of PIE ablaut patterns in verbs. So the a in these verbs must come from another source, it's not ablaut.

The clearest class 6 descendants of a PIE ablaut pattern are the ones beginning with vowels, *akaną, *alaną and *ananą, all with clear PIE cognates. In these, the laryngeal preceded the vowel, and the long ō in the past is the remnant of reduplication: stative *h2e-h2og- ~ *h2e-h2g- > *ōk-.

That leaves the class 6 verbs with a single-obstruent onset, which may or may not come from a laryngeal. CodeCat (talk) 20:42, 24 August 2016 (UTC)

That would just make these an intra-Germanic development. Positing an "a-grade ablaut" just complicates things further, especially considering PIE didn't have an independent /a/ morpheme. Even then, some of these roots did have internal laryngeals, so -aH-, > -ā- > -ō- would be expected anyway. Anglom (talk) 01:05, 25 August 2016 (UTC)