​Vowels
 
I have recently had a rather bizarre email correspondence with someone about his theory of vowels and consonants, from which belief I have been unable to shake him. The theory is that “vowels start with a fully closed, blocked glottis with the vocal folds pressed together. Consonants start with a non-closed glottis”. This, according to my correspondent, is what distinguishes the two classes of speech sound.
​
It is, of course, notoriously difficult to produce a watertight definition of vowel and consonant. If we leave aside the phonological definitions that have been advanced (that’s a different story), we are looking at the physical characteristics of the two sound classes. In Wikipedia’s article on Vowel, the lead paragraph gets off on the wrong foot: … a vowel is a sound pronounced with an open vocal tract, so that the tongue does not touch the lips, teeth, or roof of the mouth, such as the English "ah" /ɑː/ or "oh" /oʊ/.  Most phonetics textbooks mention the fact that the tongue makes contact with the upper molar teeth in non-open vowels. Additionally, palatography has shown that the tongue makes contact with the outer edges of the palate in the case of front close and front close-mid vowels. There are clear illustrations of this in Daniel Jones’ Outline of English Phonetics (see Fig 37, p. 65 for /i:/), and he also gives palatograms for “short I” and /e/. Interestingly, he says that his English /u:/ does not produce a palatogram – I think that’s true for me too, but my Cardinal  7 and 8 certainly do have tongue-palate contact at the back of the palate.

​Searching for other articles in Wikipedia that might discuss the definition of vowel, I looked at the vowel section of Articulatory Phonetics. The entire section reads as follows:
Vowels
·         Nasal vowel / Oral vowel
·         Previous Vowel / Later Vowel
·         Rounded vowel / Unrounded vowel
·         Open vowel / Close vowel
 
It seems a rather cut-price approach simply to list a few points with links to other articles. And goodness knows what the relevance of “Previous vowel/Later vowel” is. Some work is definitely needed here.

A proposal has been made by a Wikipedia editor to "deprecate" (i.e. stop using) the symbols /ᵻ, ᵿ/ in the transcription of English in Wikipedia. These are the special non-phonemic symbols invented by Clive Upton for the Oxford Dictionary of Pronunciation, to stand for "either /ɪ/ or /ə/" and "either /ʊ/ or /ə/" respectively. Wikipedia users are invited to vote on whether they want to keep or drop these symbols here. You can also see the reasons put forward for abandoning the symbols.

So far, nobody seems to want to keep them.

There is a rather strange article in the list of phonetic topics in Wikipedia, with the title Phonetic palindrome. Like many of the more oddball articles in WP it appears to have been written by a single author with little in the way of supporting scientific citations. Wikipedia itself notes “This article needs additional or better citations for verification”. The point of the article is that certain utterances are claimed to sound the same if you hear them spoken normally or hear them reversed. One simple example given is “Mom”. The article warns that there is no guarantee that the pronunciations will sound identical “because certain pronunciations can cause a shift in the articulation of the vowel, differentiating the beginning from the end in its pitch”.  We are told “In Finnish almost all palindromes are phonetic - and phonematic written”. Such statements suggest that the writer is not strong on phonetics.
The article contains quite a few examples, though regrettably most are not supported by sound files. In one case there is a reference to some work by John Oswald, who made reversed recordings of William Burroughs in the 1970’s, but if you want to find the recordings you have to look in the WP article on Palindrome in the section headed ‘In speech’. I have listened to the recordings and they just sound like unintelligible mush to me (try these examples). The Palindrome article is a much better piece of work than the Phonetic palindromes one, I think.
There is one overwhelming flaw in the whole article, however, and that is the unquestioning use of the concept of utterances “sounding the same”. Anyone who has experimented with synthetic speech knows that after you have played your bit of sound material to yourself many times, it can be clearly heard as saying what you intended it to sound like, but if you ask someone else to say what they hear, the answer is usually not what you intended. This is a form of auditory illusion. Psychologists have worked on such illusions for a long time. The “Verbal Transformation Effect” is perhaps the best-known: hearing the same thing over and over again often results in successive perception of a whole series of words, some phonetically quite dissimilar (I have been a subject in a VTE experiment, and found the effect quite disorienting). There is a Wikipedia article on Semantic satiation that describes a somewhat similar illusion.
My feeling is that if you equip yourself with some software that can play a sound recording backwards, and you then start listening to any old scrap of speech material played forwards and backwards, pretty soon you will convince yourself that you are hearing something different from what you heard at the outset. And if you are looking for bits of speech that sound the same whether played forward or backwards, you will soon convince yourself that you have found some.
Unless this article can be backed up by scientific evidence, I think it should be taken out, or moved into the Palindrome article with a lot of improvements.

The discussion referred to in my last post has developed further, as you can see here. It is now claimed that there is no proposal to remove IPA transcriptions altogether, but editors want to remove them from the "lede" and put them somewhere else. I have to say that I am still in favour of keeping things as they are.

For a long time Wikipedia has been including IPA transcriptions of words and names in the opening paragraph (see for example 'apoggiatura'), but there is currently a discussion among WP editors about a proposal to remove them as "unnecessary clutter" that is confusing to readers who don't know how to interpret the symbols. You can see the discussion going on here.

In WP, the opening paragraph is very important. It is often referred to as the "lede" (a rather pretentious spelling of "lead"), and its job is to enable the reader to judge if the article is likely to be worth reading. Some editors have suggested that if the transcriptions are retained, they should be removed from the lede and put in an "infobox", a little table usually situated at the top right of a page (see for example 'Henry Sweet').

Not surprisingly, I am strongly opposed to removing the IPA transcriptions, which I find extremely useful. Wikipedia has produced a lot of "help" material to allow users to familiarize themselves with IPA, so these transcriptions aren't inaccessible to the uninitiated (see for example here). The help pages aren't free of faults, but they are easy enough to follow.

If anyone reading this blog knows how to edit Wikipedia pages, I would urge them to add a note to the discussion to oppose the removal of IPA from WP ledes. The discussion is free for anyone to join in.

I'm pleased to say that an article I wrote for Wikipedia on the topic of the English Pronouncing Dictionary has been accepted, and can now be read here. It's only a rough first version - I hope to improve it, and I invite anyone interested to make their own improvements to it.
This comes at a time when I am quite busy with EPD-related work. I recently examined a PhD at the University of Poitiers, on research by Jeremy Castanier based on a longitudinal study of all editions of EPD from 1st to 18th, looking particularly at stress changes - a very educational experience for me. (Jeremy worked also on a bewildering range of other English pronunciation dictionaries, some of which I had never heard of). And I'm now working on a paper to be given (jointly with Jane Setter) to kick off the pre-conference event organized by the PronSIG for the 2017 IATEFL Conference at Glasgow in April. The paper will be part of a modest celebration of the centenary of the EPD, which was (you've guessed it) first published in 1917.

I notice that the Wikipedia article on Brexit doesn't yet give pronunciation guidance for this word. Is anyone out there counting how many people say /breksIt/ and how many /bregzit/? I get the feeling that the latter is gaining ground within the BBC.

There are two Wikipedia articles comparing British and American pronunciation. One is called “American and British English pronunciation differences” and the other is “Comparison of General American and Received Pronunciation”. Both of them are flawed, and have health warnings pinned to them by Wikipedia editors.The obvious question is why two separate articles should be felt necessary. The “Comparison of General American and Received Pronunciation” article is much shorter than the other, and its Talk page contains almost nothing. This makes it look like an afterthought, or some WP editor’s personal flight of fancy. WP has signalled concern over the fact that there is almost no reference to published sources in it. No reference seems to have been made to “proper” pronunciation dictionaries – everything is based on free web-based dictionaries. The writer seems not to have a clear understanding of the difference between dialect and accent (see for example the reference to “the Received Pronunciation dialect”). Although the purpose of this article would seem to be to focus on the two specific standard accents, it actually rambles on to other accents a bit (e.g. in the section on rhoticity). The glaringly obvious problem here is that the other article – the big one called “American and British English pronunciation differences” seems to be entirely based on GA and RP anyway, so the smaller article seems completely redundant.
   The big article (Comparison of General American and Received Pronunciation) seems much more substantial. Indeed, it takes a long time to get to grips with what it has to say. It starts with a comparative treatment of stress differences. The writer introduces a system of superscript and subscript letters and numbers so that s/he can indicate cases where, for example,  the pronunciation given as BrE is also the most common variant in AmE, while subscript a or b means that “the relevant unstressed vowel is also reduced to  /ə/ or /ᵻ/ in AmE or BrE, respectively” (I don’t understand the last sentence). The stress section starts with French loanwords, and we are told “French loanwords that differ in stress only are listed below”. However, examples such as ‘beret’ don’t seem to fit this – the word is given pronunciations in a note “ US /bəˈreɪ/, UK /ˈbɛreɪ/”, which in addition to the stress difference looks pretty much like a vowel difference to me. Further on, we get words cited as loanwords that only qualify as such if you can go back many centuries – e.g. ‘address’, ‘decade’. The inclusion of vowel reduction in this treatment is pretty hard to interpret – try this one (where American is initial-stressed and British is final-stressed):  ((bi)p)artisana.B1/2[nb 18] seems to cover ‘bipartisan’, ‘partisan’ and ‘artisan’, if I interpret the bracketing correctly; subscript a means there is vowel reduction, then B1/2 seems to mean that “the pronunciation given as AmE is also the most common variant in BrE” while at the same time “the AmE pronunciation of the word is a common variant in BrE”. NB18 takes us to a note that “Only the middle vowel is reduced in the BrE pronunciations”, but I can’t see any sign of vowel reduction here.
   As is well known, WP doesn’t allow material in an article that has not previously been published – such material is classed as “original research (OR)”. A WP editor has flagged an OR issue with this article, and you can see why. After a couple of sections dealing with affixes, the article moves on to segmental differences. Most of this material seems OK, though it is hard to see how it could be used easily since the table entries are given in an odd order (most WP accounts of English phonemes list them in an order that corresponds roughly to the symbols' resemblance to letters of the alphabet, so /a/ comes before /i/, /b/ before /p/ and so on.)
I have added references to the most important pronunciation dictionaries in the section “Further Reading”, but I don’t suppose anyone will take any notice, since they are not available free on line.

 
​

If you look into Wikipedia in the hope of learning about voicing and voice quality, you will find a rather confusing picture. There is a rather long and badly designed article on Phonation and this touches on most topics that one would expect to see in the description of the mechanism of vocal fold vibration. However, this article also has a paragraph on glottal consonants such as [h] and the glottal stop; this inclusion is apparently based on the unreferenced claim that “Some phoneticians have described these sounds as neither glottal nor consonantal, but instead as instances of pure phonation, at least in many European languages”. I can’t think of any reputable phonetician who would write of “pure phonation”, nor why the description should be restricted to European languages.

The article goes on to outline what it calls “Supraglottal phonation”. This is badly muddled: the list includes the terms ‘glottal’ and ‘ventricular’, which are mechanisms for producing vibration (voicing), ‘arytenoid’ (which is described as “sphincteric compression forwards and upwards”, something that makes no sense to me) and three articulations labelled ‘epiglotto-pharyngeal’, ‘raising or lowering of the entire larynx’ and ‘narrowing of the pharynx’, all of which seem to me to be modifications of phonation produced elsewhere in the larynx. There are a few useful examples of modal, breathy, faucalized, and harsh voice in Dinka. But overall, the treatment is far less transparent and comprehensive than the treatment devised by John Laver (see The Phonetic Description of Voice Quality, C.U.P., 1980).

By this stage in the article we have reached some rather esoteric material concerned with different voice qualities found in various languages. Rather abruptly we are then taken to a section on “European language examples” which is just a very simple treatment of contrastive voicing in a few languages and belongs much higher up in the article. The topic of linguistically contrastive voicing is covered by a completely different article on Voice (phonetics). The European languages section is followed by a little three-sentence section on “Phonology” and a very woolly section on ‘Pedagogy and speech pathology’; I suspect that ‘pedagogy’ here means something very different from ‘pronunciation teaching’ – there is a link to a different article on Vocal register which is very much better organised than the ‘Phonation’ article. The writer of the bit on speech pathology is clearly not aware of the vocal profiling work of Laver and co-workers.

My other major criticism concerns the sections on theories of the production of voicing, labelled ‘Myoelastic and aerodynamic theory’ and ‘Neurochronaxic theory’. Firstly, the neurochronaxic theory had been comprehensively demolished by the end of the 1950’s; it is astonishing that so many books and papers still drag out this decayed corpse for another kicking. It is as if geography textbooks routinely included a chapter on flat-earth theory, or astronomers felt the need to point out that the earth is not the centre of the universe. This should be removed from the article, and I have already told WP that I plan to do this unless there is an objection. The writer of this bit seems to think that there are two other theories, one myoelastic and the other aerodynamic, but suggests that the two might be in some way combined. As far as I know, the present-day account of how the vocal folds come to vibrate has always been known as the myoelastic-aerodynamic theory of phonation. You couldn’t just have one as a comprehensive theory of phonation. (Please correct me if I am wrong on this).

As a final headache, the topic of supralaryngeal effects on voice quality overlaps on to the topic of Articulatory settings, which I have mentioned several times recently. Some cross-referencing is clearly needed here, but I'm not sure if a separate article on Voice quality is needed if the Phonation article can be suitably reworked.
 

It's time to take another look at the issues raised by the coverage of Articulatory Settings and related matters in Wikipedia, which I first mentioned here. If you look at the very valuable comments that have been added by a number of contributors you can see that there's a lot to think about. The discussion started with the observation that WP had only a very short article on Basis of Articulation but nothing on Articulatory Settings. I added a bit of extra material to 'Basis of Articulation' and inserted a redirect so that anyone searching for 'Articulatory Setting' would be taken to 'Basis of Articulation'. 

At present I feel there are some basic issues to consider, some of them suggested by email correspondence with Piers Messum and Jose Mompean. One is that it is by no means clear that a single article should cover both 'Basis of Articulation' and 'Articulatory Setting' - there is a good case to  be made for treating them as separate subjects. Secondly, the subject of Articulatory Settings is by no means as neglected as I thought when I started looking at WP's coverage -  a lot of very interesting research has been published in the last decade or so, and this needs to be reflected in the article. Thirdly, there is no WP article at all on the very important subject of Voice Quality despite the large body of work on the subject within phonetics, so another WP article is badly needed here. It's interesting that WP has separate articles on Phonation and Voicing which for many people are the same topic.

To make a proper job of improving this complex area is going to be beyond my powers, but I am willing to make a start. Anyone can edit WP, of course, so please don't hold back if you want to contribute material. But I am starting work on a basic and brief outline of Voice Quality to submit for the WP refereeing process. Once this is accepted and added to WP, anyone interested can edit more material in, or correct errors. I will have to leave it to others to discuss whether 'Basis of Articulation' and 'Articulatory Setting' should be split into two articles.