|
|
....... |
On the Number of
Phonemes in the English language
 Counting
the number of phonemes is like counting the number of colors in a
rainbow. When we try to break up a continuous spectrum into discrete
units, we move to the realm of fuzzy logic1
by Steve Bett References
Bett- How many phonemes? 23
Brown-JSSS 27/2000/1
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| After reading the eloquent
arguments of linguist, Adam Brown [ JSSS 27/2000/1]
, I came away almost convinced that a dictionary pronunciation guide was
impossible.
Brown's intention was to explain why it was
impossible to specify the exact number of phonemes in English. He
expanded on the argument I had published in the Simpl Spel Newsletter [SSN
4/2000] . I suggested that, while 46 [21v 25c] was a good estimate
of the number of phonemes in English speech, the only number that people
would be likely to agree on would be the number of uncombined phonemes
[36] [14v 22c].1 I thought there
could be agreement on 14 pure vowels but not on the exact number
of combined vowels.
Brown's presentation reminded me of the following definition of an alphabet: ...meaningless marks arbitrarily associated with meaningless sounds. After such a definition, one might conclude that written communication is impossible. By defining the goal as a strict one-to-one correspondence to an exact number of phonemes, Brown is able to build a similar strong case against the possibility of a phonemic transcription. Almost every argument that Brown brings up is valid. A strict one-to-one correspondence requires an exact number of phonemes. If we cannot agree on the exact number of phonemes then we cannot have a corresponding alphabet or phonemic transcription. If we set the level of precision high enough, a one to one correspondence between graphemes and phonemes can be shown to be impossible to attain. In a [04June2000] letter written to the saundspel phonology forum, Michael Avinor put it this way, "Speech is an analog signal and writing is a digital signal. To talk about a phoneme we have to cut up continuous speech into discrete units. Digitizing speech can preserve only a limited part of the speech information." Even more information is lost when speech is visualized or represented graphically. Nevertheless, the fragment of the original that remains can be enough to accurately convey information. Sounds are not the only perceptions with fuzzy boundaries Sounds are not the only things in our perceptual world with fuzzy boundaries. Has anyone ever claimed that a name for a discrete segment of the sound spectrum was any more exact than a color name for a desecrate segment of the color spectrum? The flaw in the Brown's argument is the implied insistence on a high level of precision. If we raise the bar of precision high enough, then most ideals can be characterized as unobtainable and unrealistic. We cannot see a particular color, e.g., blue, any more than we can hear a particular phoneme. This does not mean we cannot discriminate or sort blue and yellow objects. We can be presented with an instance and then asked to judge whether or not it a member of a category or class of 'blue' things. As we get near the boundaries [e.g., Is a green ball sorted into the pile of blue objects or yellow objects?], the judgments become more uncertain. However, there are some modal or mid range instances that nearly everyone will agree is a particular primary color. There are even instances that nearly everyone will agree is blue-green. When the blends become more complex, however, agreement becomes harder to achieve. The Phoneme - fuzzy by definition: a range of sounds treated as identical The same is true for the abstract sounds we call phonemes. Phonemes have been defined as the smallest unit of sound capable of changing the meaning of a word. The substituion of b for p in [pit] changes the meaning. Therefore [b] is a phoneme and the p:b distinction is phonemic. [pit : bit] are called minimal pairs. The substitution of @ for i in exact changes the meaning. A phoneme is not really a unit in the sense of a single sound. It is a range of related but acoustically distinct sounds treated as a unit -- treated as if they are the same sound. What we hear are sounds or phones. A phoneme is an abstraction or interpretation. The same person will pronounce the same vowel in acoustically different ways in association with different consonants. People from different speech communities will rarely pronounce the same vowel the same way. Three people will have three different concepts
as to what phones [physical sounds] should be accepted or rejected as members
of a given phoneme. However, there will be an area where the three
concepts overlap - an area of agreement.
100% agreement is possible for 14 clear
instances
Any time you try to break up a continuum or spectrum into discrete units, there will be problems at the boundaries. Not everyone will slice the continuum at precisely the same point. However, just as it is possible to achieve nearly 100% agreement on instances of the primary colors, it is possible to get nearly 100% agreement on instances of the primary or uncombined phonemes. Words are not pronounced the same in different dialects of English which means that a phonemic representation of one dialect may not correspond to the the speech in a different region. A phonemic script always presumes a base dialect and such alphabetical writing systems developed for one dialect will not always be a reliable guide to the pronunciation of another dialect. Thus some people will have to learn a spelling dialect in addition to their local dialect. This situation is also true in Spain and Italy. The base dialect, Castilian, for instance, does not always correspond to the local dialect. Not everyone will agree where i sound stops and e sound begins. The disagreements will increase when the vowel is unstressed. The last phoneme in <vegetable> can be transcribed
as vejt@bl, vejt@b@l,
or vejt@bUl.
The first phoneme in <because> can be transcribed as becoz, bicoz,
or b Breaking up the vowel sound continuum into discrete units is analogous to breaking up the color spectrum into discrete colors. Adding gray into to a color mix could be an analog of removing stress in speech. Gray [or the lost of brightness and contrast] reduces the ability to discriminate colors just as the loss of stress reduces the ability to discriminate vowels. Few people can detect the difference between three pronunciations of [pencil], e.g., pensil, pens@l, and pensl. Yet most will be annoyed by the written substitution of pens&l, pencl, pensl, or pensal for [pencil]. The objection to the loss of familiar "eye rhymes," however, is another issue. When it comes to blends, one may have to make
some arbitrary decisions. Sweet and Jones, for instance, decided
not to include triphthongs [or trigraphs] in their list of 46 phonemes.
Many phonemic systems, such as Unifon and Truespel, arbitrarily limit
the number of phonograms to 40.
Linguists, such as Daniel Jones, broke the sound spectrum down into three parallel segments [see chart]. The parallel segments have been called tense, lax, and rounded. The same sound may have three different expressions depending on muscle tension and rounding. The other dimension of speech sound goes by several names. One reference is to the jaw position: open, half open, or closed. Tongue position during vocalization [front and back] are also used. Typically four levels of openness are isolated Finally, there are glided blends of two sounds [diphthongs] that start in one position or cell of the chart below and end up in another. ai, for instance, is a combination of @ or a: and i or i: [ai = aah+ee] One phono-graphic representation cannot cover all the dialects of English This is a different issue than the one just discussed. It is not a question of not being able to recognize clear instances of a phoneme. Rather it is a case of which phonemes should be used with particular words. If the symbol represents a speech sound then when that sound changes the symbol has to change. This is true unless there are some symbols that merge two sounds. [o] for instance, could represent both the short /aa/ and the short awe. If one wanted to be specific about the sound, then another symbol coud be employed, e.g. au or Q, as in "ai baut aul our aarmy cots." This would be an easy convention to add to the truespel notation [www.truespel.com] since the o is not used alone. How do you spell a word that has three different pronunciations depending on the dialect of English someone speaks. Phonemic notations require a base pronunciation. The following tables contrast two regional
dialects of English.
ASCII
http://www.hpl.hp.com/personal/Evan_Kirshenbaum/IPA/english.html Not all digraphs are true diphthongs Not all digraphs are blends or glides. ae, ei, ou and aw are not necessarily blends as would be suggested by their common representation as digraphs in some notations. One can produce sounds which will be interpreted as members of these phonemes without moving one's mouth or position of one's tongue. [ae, ei, ou and aw] pronounced either as pure vowels or as diphthongs will be understood as allophones of the phonemes represented by these digraphs. Were there an abundance of vowel shapes or graphemes, these four sounds typically referenced by a digraph would have been represented with one symbol - as they are in the IPA. However, in the case of /ou/ there would be dialects where /@u/ would be a more accurate narrow transcription of regional speech. As it happens, the pure vowel ei is almost indistinguishable from the blend of e [eh] + i [ee]. The pure vowel aw [as in law] is pronounced in various ways in different dialects. The diphthongal pronunciation [ aa+U] can be understood in context. The pure vowel O or Spanish O can be distinguished from the dipthong /ou/ [awe+oo] but both can be readily understood. The combination, @U, would be closer to the big O sound but this is not a common representation in TS. The primary phonemes that are obscured in the traditional orthography can be problematic. [ ^/@, 3:r/@r, dh/th, ng/n-g ] . Some transcription systems see no need to reveal obscured phonemes. As Brown observes, these obscured phonemes usually carry a low functional load. Notations and Dialects
With a phonetic notation, it is much easier to write dialect than with traditional unsystematic spelling. The goal of a new script such as Spanglish, however, is to ignore dialect differences - to be pan dialect or diaphonic. In Spanglish, as shown above, there are usually
two ways to spell a sound depending on whether it is stressed or unstressed..
hurrder shows stress, herder does not, hudda corresponds to IPA /'hLd The basic difference in the minimal pairs [^/@] , [3:r/@r] is stress. If the orthography has another mechanism for representing primary stress [as with Truespel], the need to distinguish these vowels with different letter shapes would be less urgent. The minimal pair in the/thug or thy/thigh can be easily distinguished but there are only a handful of words whose meaning would change if these consonants were mispronounced, e.g., voiced instead of left unvoiced. [see voicing] The same is true for the minimal pair that distinguishes schwa [@] from [^]. @gact igstract Minor differences in pronunciation Brown makes much of regional differences in the pronunciation of the vowel in some words. The pronunciation may be evenly split between two slightly different sounds. [dog], for instance, can be dawg or daag. [bath] can be baath or baeth. These are real problems but they need not lead to paralysis. In Spanglish, dog would be written [dog] and
pronounced [dawg]. The o is very short and closer to [awe]
than [ah]. This is close enough for speech recognition and close
enough for a phonemic orthography. To express the pronunciation /da:g/,
Spanglish uses daag or dogg. To transcribed southern drawl, dog
would be written <dawg>.
Finer distinctions between dialects are always possible IPA limits the number of vowels to 21 and lists 26 consonants. Romik arbitrarily limits the number of phonemes to 25 vowels and 25 consonants. It is almost always possible to make finer distinctions even when allowable distinctions are limited to minimal pairs. The generic saundspel egroup notation, Romik, includes a purposely ambiguous a, o, and 'r. These can be specified for particular regional dialects such as [GA] but they can also be left broad and vague in cases where the pronunciation dictionary gives two nearly equally common pronunciations or when the pronunciation dictionaries for [GA] and [RP] disagree. It is not unusual to have two correct pronunciations of one spelling and both can be understood. The goal is an orthography almost as good as existing pronunciation guides Although Brown overstates his case, he makes some good observations. There is no point in striving for a transcription system that is more precise than existing pronunciation guides. Pronunciation guides will not always be based on the same number of phonemes and will not always be in 100% agreement. A pronunciation guide for British English will be slightly different from a pronunciation guide for General American. The differences are addressed in Kelley's vowel chart for Englik and have been discussed on the saundspel phonology forum. However, all pronunciation guides will be more alphabetic and phonemic than the traditional writing system. The goal is to get closer to the ideal, not necessarily to achieve it. The goal in many minds is simply a writing system that is as consistent and alphabetic as Italian or Spanish. This goal clearly falls short of a "strict one to one correspondence." Most attempts to regularize spelling will usually bring the writing system closer to the ideal. There are two types of goals, one is a simple expression of the desired direction the other involves the expectation that the goal will be achieved. Brown argues that since the goal of a perfect onetime correspondence between phonemes and graphemes is unobtainable the goal should be abandoned. If there are those who think that we can actually count the number of colors in a rainbow, then the point is well taken. However, the majority of orthographies are simply trying to find a better way to do what has already to been done in other languages. A majority would be satisfied if English could
be written in as transparent a way as Italian.
Are there 48 sounds in English as the chart indicates? Certainly there are clear instances of each sound category but this does not mean that they all have to be unique phonograms. Any number greater than 36 is somewhat arbitrary although a good case can be made for including the combined phonemes ch, j, as unique phonograms. Most systems include the diphthongs: ai, ou, and oi. Many refer to /ai/ as a long vowel as if could be produced without moving one's tongue. 40 phonograms is the most popular number. Note that the two vowels in [herder] are listed as pure vowels. The basic difference between the two is stress. M-W represents this sound as ['&] a stressed mid lax vowel. [er/urr] is written as if it were an r-combination but it is a single sound /3:/. The other r-combinations cannot be articulated as a single continuous sound.
In GA, the letter o in most words is pronounced the same as [aa]. Pronunciation guides for American English will often use the same symbol for [o] and [aa]. As shown above, <bother> is pronounced /baather/ in GA. The pan-dialect solution is to include all the phonemes an use a hybrid base dialect. bother would have a British flavor in the first syllable and a rhotic American flavor in the second. The pan-dialect solution is to keep the short o but allow two interpretations. In the writing system, o would be a little ambiguous. In the pronunciation guide another set of symbols [e.g. q for the turned a and Q for the British short o] would be used. In Spanglish o=awe and and oCC = aa. doter = daughter and otter =aater Conclusion We cannot count the number of vocalic phonemes in English speech any better than we can count the number of colors in a rainbow. However, just as we can identify the primary colors in the rainbow, we can identify 36 clear instances of the primary or uncombined phonemes in speech [14 pure vowels, 22 pure consonants]. As we try to make finer distinctions, unanimity of opinion declines. There will never be much agreement on the exact number of combined vowels. Most people will want unique phonograms for [ch-tsh], [j-dzh], and [ai] bringing the total to 39. [oi], [au] and a few r- combinations should also be singled out. [46] is a good number of pure and combined phonemes. This would be enough to represent the sounds in both British and American dialects. [IPA-40 chart] As the chart above
shows, we can get by with 42. There will be words that continue to be pronounced uniquely in a particular dialect. Thus a transcription system based on General American with allowances for Southern British may not always accurately represent the pronunciation of some words in these dialects. It is not that these dialects have any more phonemes [although this is a possibility], they just apply them differently in a few words. [Review dialect transcription] The goal is to develop a workable writing system for English that is as good as the Italian writing system is for the Italian language. Unlike the goal of perfect one-to-one correspondence between graphemes and phonemes, this goal is attainable. What prevents its realization is not the elusive goal but the fact that any consistent system will respell 60% of the words in English and that most of these respellings look odd to those adept in the traditional writing system. Some respellings will "offend the eye." For those who have acquired a high level of word pattern recognition, respelling seems to obliterate certain distinctions isolated by heterographic homophones [know, no] [dough, doe]. One symbol per sound can remain the stated goal without the expectation that it is the kind of ideal that can ever be fully attained. There is certainly no need to abandon this idealistic goal at this point. It clearly defines the correct direction. When the English writing system becomes as good as one of the systems used as a pronunciation guide in a dictionary, then the quest can be abandoned. Pronunciation guide spelling that is near 100% predictable is about as close as one can expect to get. Beyond this we quickly reach a point of diminishing returns. The goal is not to be better than the dictionary pronunciation guide but to approximate it with a practical everyday writing system devoid of ukknsukpported special characters and complicated diacritics. We are trying to come up with the best possible visual representation of the abstract phonemes that people have in their heads. We are trying to achieve a system or representation that is nearly isomorphic with the phonological structure of English speech. We will never reach this goal. Fortunately, a system that is less than ideal will be "good enough". A writing system for English that is as good as the writing systems for Italian and Spanish will be fine. There is no perfect perfect graphic representation of speech sounds. Since the writing system is not used to capture subtle differences between different dialects, the system does not have to be as detailed as IPA. As good as Spanish is quite adequate for English. In linguistics we are working in a realm of fuzzy logic not aristotelian logic where everything is either black or white . . . true or false. Brown points out all of the limitations of phonemic spelling and then concludes that since the goals is elusive it should be abandoned. In building a better system for a broad transcription of English, there is a point of diminishing returns. This point will be reached long before we have to become concerned about the precision of phonemes or the suitability of a particular base dialect. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Definitions
fuzzy logic - in classical logic everything was black or white,
true or false. fuzzy logic recognizes a middle ground, e.g., usually
true. Fuzzy logic is a superset of conventional (Boolean) logic that has
been extended to handle the concept of partial truth -- truth values between
"completely true" and "completely false". It was introduced by Dr.
Lotfi Zadeh of UC/Berkeley in the 1960's as a means to model the uncertainty
of natural language. [more]
phoneme - a difference in sound that makes a difference in meaning - a range of sounds treated as the same sound. A phoneme is abstract concept or category - you cannot see, touch, or hear a phoneme but you can point to instances. A phoneme is not one sound but a family of sounds, especially when more than one speaker is involved. A phoneme is an area. All instances in that area are referred to as allophones or diaphones. Phonemes are language specific.
Where English speakers distinguishes two phonemes
Phonemes are called the smallest unit of meaningful sound in a language. An orthography where the letters represent sounds is called phonemic.
To answer the question, look
it up in a pronunciation guide and count the phonograms.
How many phonemes in the word thorough? The dictionary says qer'u. Looks like 5. This is a little tougher question because it is uncertain if ['u] is one or two phonemes. Unifon TerO Spanglish thero Truespel thheroe. Xenglik TerO. Unifon's one sound per symbol design suggests that it is probably the best transcription system for easy counting.
Notes: Writing in 1891 E.V. Graff presented a phonetic alphabet for 37 elementary sounds. This is the same as the one above except for the addition of hw as in when and where. References Bett, Steve. 2000. Can we pin down the number of phonemes in English: Simpl Speling Newsletter, April, 2000, p. 7. [phon-inv] [old version at fortunecity] Bett, Steve. 2001. On the number of phonemes in English. JSSS, fall, 2001 Brown, Adam. 2000. The number of phonemes in English: not a simple answer to a simple question. JSSS 27/2000/1 pp. 11-13 Brown, Roger. 1989. Words and Things.
Glencoe Free Press.
Wells, John. SAMPA- speech assessment methods phonetic alphabet, 1987 http://www.phon.ucl.ac.uk/home/sampa/home.htm How many phonemes? 2 3 long version of the published article. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Expanded References: Aaron, P.G. and R.M. Joshi. 1989. Reading and Writing in Different Orthographic Systems. Kluwer Academic Publishers in Cooperation with NATO Scientific Affairs Division.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 14 pages |
© 2000
BETA
Information Design
|
Above
are the 21vowels isolated by D. Jones and used by most scholars
[e.g., 
coz. The less stress
or smaller the discrete unit, the less agreement there will be.
40
40
42
46
r'vet
iz mai 'feiv
According
to the chart and the two letter representation, [ei] looks like a diphthong.
However, [ei] is not a diphthong in the same sense as ai,
oi
and au. [ei] can be pronounced as a single unobstructed
and unaltered tone. It can be a pure vowel. Phonemically, however,
it would not be distinguished from the e+i blend in the word vein.
[ow] is another case of a possible pure vowel being represented
as a diphthong. A majority of English speakers, however, probably pronounce
[ow] as a diphthong. If the orthographer's goal is to minimize the
number of symbols, he might treat both [ei] and [ow] as diphthongs.
The
exact identity of unstressed vowels will continue to be subject to different
interpretations. People will continue to argue about such things
as [1] what would be the best way to represent unstressed sounds
[to schwa or not to schwa...]. [2] should the redundant kqx be in
the alphabet, [3] should both the voiced and unvoiced [th] be included,
etc.
