The notion of syllable has been discussed for long time under different theoretical approaches. A linguistic unit such as syllable is necessary to give theoretical support to practically all linguistic theory. Some researchers made it an idealized linguistic unit. But syllable is quite real. The problem is where and how to find it. Most phoneticians demonstrated that the syllable is an aerodinamic effect upon the airstream in speech (cf. Catford, Ladefoged, etc.). Unfortunately very few research have been done in this direction. If you tap your back when saying a long aaaaa in continuous movement, the interruption of the air flow from the lugs will give you controlled syllables (long vs short). Syllable is an empathic way to interpret what we hear: if someone speaks not following the general patterns of the language, the hearer will feel confused and annoyed. Moreove, the kids play with the syllables, we use the idea of syllable to make rimes in poetry. We sing and we talk following certain rhythms. Without the syllable the language we speak would be impossible. However, dont ask Praat to give you the segmentation of speech in syllable. Try to hear and principally to feel the syllable. The linguistic theories simply follow what the speakers feel and think about their languages...

The mental reality of syllable has been discussed in heat for many years.

As for general linguistic and phonological discussions the following references may be interesting for you:

Ladefoged, P. (1967). Three areas of experimental phonetics: Oxford University Press.

Hoard, J. E. (1971). Aspiration, Tenseness, and Syllabication in English. Language, 47(1), 133-140.

Bailey, C.-J. N. (1978). Gradience in English Syllabification and a Revised Concept of Unmarked Syllabization. Bloomington: Indiana University Linguistics Club.Bailey, C.-J. N. (1978). Gradience in English Syllabification and a Revised Concept of Unmarked Syllabization. Bloomington: Indiana University Linguistics Club.

Roach, P. (1982). On the distinction between ‘stress-timed’and ‘syllable-timed’languages. Linguistic controversies, 73-79.

Gussenhoven, C. (1986). English plosive allophones and ambisyllabicity. Grammar.

As for phonetic and psycolinguistic discussions the following references may be interesting for you:

Dauer, R. M. (1983). Stress-timing and syllable-timing reanalyzed. Journal of Phonetics.

Meyer, A. S. (1990). The time course of phonological encoding in language production: The encoding of successive syllables of a word. Journal of Memory and Language, 29(5), 524-545.

Zwitserlood, P., Schriefers, H., Lahir, A., & van Donsellar, W. (1993). The role of syllables in the perception of spoken Dutch. Journal of Experimental Psvchology· Learning,Memo/T, and Cognition, 19, 260-271.

Levelt, W. J. M., & Wheeldon, L. (1994). Do speakers have access to a mental syllabary? Cognition, 50(1-3), 239-269.

Cutler, A., Otake, T., & McQueen, J. M. (2009). Vowel devoicing and the perception of spoken Japanese words. The Journal of the Acoustical Society of America, 125, 1693.

Hanulíková, A., Mitterer, H., & McQueen, J. M. (2011). Effects of first and second language on segmentation of non-native speech. Bilingualism: Language and Cognition, 14(04), 506-521.

Verdonschot, R. G., Nakayama, M., Zhang, Q., Tamaoka, K., & Schiller, N. O. (2013). The Proximate Phonological Unit of Chinese-English Bilinguals: Proficiency Matters. PLoS ONE, 8(4), e61454.

Rietveld, A. C. M., & Schiller, N. O. (2014). Phonetic Accounts of Timed Responses in Syllable Monitoring Experiments. In J. Caspers, Y. Chen, W. Heeren, J. Pacilly, N. O. Schiller & E. v. Zanten (Eds.), Above and Beyond the Segments. Experimental linguistics and phonetics.

Unfortunately, I don't know German. But in some linguistic environments in English, the concept of syllable may be useful. Such environments may include, for instance, the rules of stress in phonology. Without syllable count and/or determination, especially for SLA learners, the stress on the penultimate syllable will not be rightly distributed and learned as in e/co/NO/mic. Moreover, in morphology prefixation and suffixation occupy a single syllable. Prefixation, in particular, occupies a syllable which usually does not receive stress in English. However, all this is done at the cognitive unconscious level by native speakers who will not need the help of the syllable in naturally-occurring or carefully constructed discourses.

Klaus:

Syllables are a essential phonological unit in early colonial central Mexican Nahuatl and are essential, as Zouheir comments regarding English, for understanding lexical morphology. For example, in this Mesoamerican language, syllables must belong to one of four types: V, CV, VC, and CVC (where V is a vowel and C a consonant or semiconsonant). Unlike English, there cannot be more than two consecutive consonants in the interior of a word, and there cannot be two consecutive consonants at the beginning or the end of a word. To avoid violating these rules, the vowel /i/ is inserted between consonants. If one does not consider these rules, morphological analysis is not possible, as these /i/ vowels have a phonic rather than a phonemic value, that is, they do not contribute to the semantic value of a word, and in certain contexts they are dropped. Thus the concept of the syllable is a very useful one for anybody doing lexical analysis of Nahuatl discourse.

For reading pre-Hispanic Maya writing the concept of the syllable is also essential, since this system combines graphs representing syllables with others representing morphemes or ideas.

Dear Junru Wu (what its the first name?),

thank you very much for all the references. I will try to read all. I agree, when you say: "The mental reality of syllable has been discussed in heat for many years." But I don't see a trustable result.

Dear Zouheir,

you say we need the category of a SYLLABLE for the rules of stress in phonology. But the stress is always on a vowel. We can make rules on which one the stress is.

Regarding morphology I understand that there are morphemes in most languages. But they are not always identical with syllables. In English we have "(He) doesn't (come)." or "(He) goes." where you find one syllable for two morphemes. And what about "coming"? Where is the syllable boundary? Is it identical with the morpheme boundary? Doesn't the category SYLLABLE make all these things more complicated?

Dear Robert,

couldn't we say there are Maya signs for certain sequences of sounds? And if we wouldn't need the 'syllable' we wouldn't need the V, CV, VC, and CVC either.

The only argument which convinces me is the epenthesis-argument.

Dear Robert,

you wrote: „Syllables are important for emphasizing points within a word“. Isn’t it sufficient to look at which vowel you emphasize?

Furthermore it is not easy for me to understand how syllables help ro recognize denotation and connotation.

I tried to do some corrections to my contribution. But it didn't work. Instead of "David" I wrote "Robert" in my last but one contribution.

Am I the only one who has problems with the research gate program?

Klaus: a syllable may be thought of as a phonological unit, independent of meaning, minimally a single vowel and possible including another vowel (creating a glide) and one or more consonants before the nucleus (onset) and/or after the nucleus (coda); thus it is a unit of sound in verbal discourse. A morpheme is a unit of sound plus meaning, and may have one or more syllables, or none (for example, the plural suffix -s in English, which has no vowel). Syllable boundaries and morpheme boundaries may or may not coincide. If we look at the word 'microscopic,' for example, and mark the boundaries with periods, the syllable boundaries are: mi.cro.sco.pic, while the morpheme boundaries are: micro.scop.ic. In this example one of the three syllable boundaries coincides with one of the two morpheme boundaries. (There may be some discussion about the placement of syllable boundaries depending on one's theoretical approach.) So I find both 'syllable' and 'morpheme' to be useful concepts for thinking about phonetics and semantics, respectively. I don't understand why we would want to remove the concept of 'syllable' from our conceptual toolbox.

Klaus: when the software won't work I refresh (reload) the page and click on the "Edit/Delete" triangle in the upper right corner of the message box. That does the trick for me.

As for Maya epigraphy, yes, we could call the syllables "sequences of sounds", but I find the 'syllable' concept much more precise. Different languages have distinct rules for syllable formation, and these rules are taken into account when parsing words into their constituent morphemes, as in the example of the 'filler' vowels /i/ in Nahuatl. The concept of syllable lets me see what is happening in lexical morphology.

People who look at the aesthetic aspects of verbal discourse in literary analysis would have a hard time doing their job without the concept of syllable, I would imagine, since the phonic values may be even more relevant than semantics, especially regarding rhythm.

Dear Klaus

Quirk et al (1985) devoted an Appendix (p. 1587) to "stress, rhythm and intonation" in their now classic, A Comprehensive Grammar of the English Language. On p. 1591, they repeated many times that "the stress falls on the syllable" not the vowel nor the consonant. A look at my Daniel Jones English Pronouncing Dictionary revealed that the syllable on which the stress falls almost always starts with a consonant followed by by a vowel such as in in/duce, flac/ci/di/ty, rui/na/tion, etc. (I chose these words randomly).  

Regarding morphology, I don't think I said that all morphemes are identical with syllables. I selected prefixes and suffixes, which occupy a syllable in word initial (e.g. UN/caring) or word end (child/LESS). As to COMING, COME is a one-syllable word to which is added the aspectual suffix (ING). The syllable boundary is between the root (COME) and bound (ING) morphemes. Obviously, syllable is identical with morpheme if the latter is a one-syllable word or morpheme. In multi-syllable root morphemes, the syllable does not overlap with the syllable as in UNpredictable, where only the prefix coincides with a syllable, predictable being a three-syllable morpheme. Therefore, I don't think that syllable is making our life difficult.  

David,

thanks for the trick with the sortware. But even that didn’t work. Anyway…

„a syllable may be thought of as a phonological unit, independent of meaning“. My problem is, why. Why phonological units without meaning? Maybe independent of any sense outside the endless plays of ‚syllable phonology’.

And (if we want syllables): Why not „mi.cros.co.pic“?

„I don't understand why we would want to remove the concept of 'syllable' from our conceptual toolbox.“ That’s my question: Why not?

„but I find the 'syllable' concept much more precise. “ What is precise with the syllable concept (esp. if we even don’t know ist boundaries)?

„Different languages have distinct rules for syllable formation, and these rules are taken into account when parsing words into their constituent morphemes“ How ist hat possible if the morpheme boundaries are not the same as the syllable boundaries?

„the aesthetic aspects of verbal discourse“ – what is aesthetic with a syllable? The syllable of maximum length in German is „strampfst’s“ (CCCVCCCCCC) – what’s aesthetic with that?

Dear Zouheir:

„Quirk et al (1985) devoted an Appendix (p. 1587) to "stress, rhythm and intonation" in their now classic, A Comprehensive Grammar of the English Language. On p. 1591, they repeated many times that "the stress falls on the syllable" not the vowel nor the consonant.“ Not the vowel? Why not? Maybe they are not right?

„A look at my Daniel Jones English Pronouncing Dictionary revealed that the syllable on which the stress falls almost always starts with a consonant followed by by a vowel such as in in/duce, flac/ci/di/ty, rui/na/tion, etc. (I chose these words randomly).“ But it doesn’t fall on the consonant, it falls on the vowel.

Suffixes in German are often not identical with syllables: e.g. „Zei-tung“ (newspaper), the suffix is „–ung“. German phonologists would analyze the syllables of „coming“ as „co.ming“ or „com-ming“ (before or in the mid oft he „m“), never after the „m“. Root and bound are no criteria for German phonologists. The pronounciation ist he only criterium for them.

„Obviously, syllable is identical with morpheme if the latter is a one-syllable word or morpheme. In multi-syllable root morphemes, the syllable does not overlap with the syllable as in UNpredictable, where only the prefix coincides with a syllable, predictable being a three-syllable morpheme.“ In this case it should be sufficient to have two concepts (word and morpheme) instead of three.

Klaus, here are my answers (A) to your questions (Q):

Q - Why phonological units without meaning? Maybe independent of any sense outside the endless plays of ‚syllable phonology’

A - A syllable, as a phonological unit, has no intrinsic meaning; if it did, I would reclassify it as a monosyllabic morpheme.

Q - And (if we want syllables): Why not „mi.cros.co.pic“?

A - That would be fine with me, if you can justify it phonologically. This is why I included the caveat "(There may be some discussion about the placement of syllable boundaries depending on one's theoretical approach)" in my last post.

Q - „but I find the 'syllable' concept much more precise. “What is precise with the syllable concept (esp. if we even don’t know ist boundaries)?

A - English is sticky, with its consonant clusters. In Nahuatl the boundaries are not so ambiguous. Perhaps this matter should be discussed as it applies to individual languages! The questions and answers would probably vary from one language to another. I used English examples out of deference to the lingua franca we are using on this thread.

Q - „Different languages have distinct rules for syllable formation, and these rules are taken into account when parsing words into their constituent morphemes“ How ist hat possible if the morpheme boundaries are not the same as the syllable boundaries?

A - I tried to make that clear with my example of the "filler" vowels /i/ in Nahuatl (Richard Andrews calls them "supportive" vowels, while Frances Karttunen prefers to speak of "weak" vowels). I see now I was leaving out part of the explanation. Knowing about the rules for syllable formation and other aspects of Nahuatl phonology enables me to distinguish the filler vowels from the same vowel /i/ when it is used as part of a morpheme. The filler /i/ in certain morphophonological contexts is ellided, where an /i/ that forms part of a morpheme is not. If I let go of the concept of the syllable, and with it the rules of syllable formation, I will have a hard time distinguishing between the two sorts of vowels /i/ and their different morphophonological interactions on the borders between morphemes.

Q - „the aesthetic aspects of verbal discourse“ – what is aesthetic with a syllable? The syllable of maximum length in German is „strampfst’s“ (CCCVCCCCCC) – what’s aesthetic with that?

A - Again, perhaps this discussion should be carried out on a language-to-language basis. Germanic language consonant clusters can get quite thick, as you show, although I would not say that limits the aesthetic possibilities of these Indo-European tongues, just that they have their own particular aesthetic possibilities, like any other language. I mentioned rhthym, which is an important formal aspect in the aesthetics of speech, particularly in poetry and song, and a basic rhythmic unit is the syllable, marked by the consonants cutting the vocal stream into phonic parts which can be strung together in different ways to ajust the meter.

And now I have a question for you, Klaus: Why do you "think within fluently spoken language we can’t hear syllables," as you state in the introductory text to your question? Do you think this applies to all languages? I don't understand or speak German, but I am sure that I can hear syllables in English, Castilian, French, Otomi and Nahuatl. As we both have pointed out, their borders may be blurry in some cases, but I don't think that blurriness negates their existence. I do have trouble hearing boundaries between words sometimes. Perhaps the concept 'word' is on less solid linguistic ground than the concept 'syllable.' Still, I wouldn't discard it, because it is also useful to me for practical purposes (especially writing) and for metalinguistic communication with other people.

One final question, Klaus. In the introduction to your question, you say "But I think another description might be possible." Can you please provide that description so that the rest of us might try to understand what you are getting at? Perhaps this might help us move this interesting discussion forward.

Utility of the term "syllable" depends fully on where one is standing and whether comparisons are made.

If one is comparing and contrasting languages (interlanguage), which multilingual students do all the time, there is not the sort of focused intensity on syllables as in those who are teaching in a mono-lingual environment (intralanguage), like regular English classrooms in the UK or USA.  

Both (interlanguage and Intralanguage situations)  focus however on the rhythm and tone of the spoken language.  In some languages, the rhythm leads to more focus on the syllables. 

Here are my thoughts on the syllable:

It is not a universal unit. It is a derived unit that can emerge with different structures in different languages. In some languages (e.g. French) it is relative easy to determine boundaries between syllables. In others, like English, it can be exceedingly difficult, leading to proposals such as 'ambi-syllabicity' (Kahn 1976). The issue of boundaries suggests that languages differ in a more significant way than just answers to questions like: What languages allow 'branching onsets' for 'complex codas' etc.

I very much agree with Kevin and Geoffrey.

Dear David,

I am just going to change my temporary residence from Macedonia to Bavaria. So it is difficult for me to answer all questions thoroughly. But I’ll try some short statements:

„A syllable, as a phonological unit, has no intrinsic meaning“ – I know. But why such a big discourse about things which have no meaning? (Especially as ‚language = meaning’)

„That would be fine with me, if you can justify it phonologically. This is why I included the caveat“ – why justify s.th. phonologically which phonology first seems to need for its own justification?

Your caveat: I did recognize it, and I agree.

„The questions and answers would probably vary from one language to another.“ I agree.

„I tried to make that clear with my example of the "filler" vowels /i/ in Nahuatl“ – As I wrote before the epenthesis-argument does convince me.

„Again, perhaps this discussion should be carried out on a language-to-language basis.“ – A agree with you.

„these Indo-European tongues, just that they have their own particular aesthetic possibilities“ – I agree with you. The aesthetic measures of German are counted by the number of stresses, not by the length or heaviness and composition of syllables.

„And now I have a question for you, Klaus: Why do you "think within fluently spoken language we can’t hear syllables," as you state in the introductory text to your question? Do you think this applies to all languages?“ – No. I thought of languages like German. There are some (very few) German researchers who prefer the idea that some languages (like German) are word languages, phonologically spoken, and that one should prefer to analyze the word structure instead of the syllable structure when investigating these languages.

„One final question, Klaus. In the introduction to your question, you say "But I think another description might be possible." Can you please provide that description so that the rest of us might try to understand what you are getting at?“ I have two ideas. The first is, concerning certain languages like German and maybe English etc., to concentrate merely on the vowels, not on syllables as little systems and their extreme complicated structures, which cannot be prooved sufficiently. The second is to distinguish between languages with clear syllable structure and languages like German which have a word structure. So we would have syllable languages and word languages. But even concerning syllable languages I would say the syllable should not play the rule of „the“ central phonological category, because it has no meaning, and language in the first place deals with meaning.

Thank you for your answers, Klaus. Now I can see more clearly what you are thinking. This discussion is very interesting to me, since I rely heavily on linguistic categories, from allophones to discourse (including phonemes, syllables, morphemes, and words), plus the extralinguistic categories of ideas and their visual expression (semasiographs), in my work with pre-Hispanic and early colonial Mesoamerican writing systems. These systems are mixed, expressing all of these categories by means of simple and compound pictorial graphs, and frequently include the representation of syllables to construct graphs that express words and thus convey meaning. The Western categories of "writing" and "visual arts" fuse in this complex form of pictorial writing that functioned well in ancient Mesoamerican multilingual society. The reason it functioned well is because its fundamental base is semasiography (the graphic expression of ideas), although the system permits some glottography (the graphic expression of linguistic elements: phonemes, syllables, morphemes, and words, through homophonic word play, like rebus writing). I am speaking of central Mexican pictorial writing; the Maya script tends more toward glottography, while including some semasiography.

It looks to me like the consonant clusters in Germanic languages weaken syllabic structure to the point where you and others feel the need to express concern about the validity of the concept. Perhaps the historical influence of Latin grammar in the description of languages from other families is part of the problem. This has certainly been the case with Mesoamerican languages, whose first grammars were written by friars under the influence of descriptions of Latin, and also of Nebrija's Castilian grammar of 1492, which follows the same patterns. The Mesoamerican language grammars are surprisingly profound and insightful, but they never completely succeeded in describing native tongues in their own terms.

If language "in the first place deals with meaning," spoken language expresses meaning through sound, so there is a complex interplay between semantic content and phonic form. Some people focus more on one aspect, some on another, some on the interplay, and others on the complex whole. As Kevin pointed out, one's perspective may determine what one tends to focus on.

Thank you, David, for this very interesting answer. What you work on is really fascinating.

You are welcome. Klaus. I just went back and expanded it a little, because I realized that I hadn't explained why Mesoamerican pictorial writing worked well in a multilingual society. In our times we are moving toward a greater use of semasiography, I think for the same reason (the need for communication among speakers of diverse languages), as can be seen in highway signs, electronic device controls, and graphic user interfaces for computers. In semasiography, of course, semantic content totally trumps phonic form.

The notion of syllable has been discussed for long time under different theoretical approaches. A linguistic unit such as syllable is necessary to give theoretical support to practically all linguistic theory. Some researchers made it an idealized linguistic unit. But syllable is quite real. The problem is where and how to find it. Most phoneticians demonstrated that the syllable is an aerodinamic effect upon the airstream in speech (cf. Catford, Ladefoged, etc.). Unfortunately very few research have been done in this direction. If you tap your back when saying a long aaaaa in continuous movement, the interruption of the air flow from the lugs will give you controlled syllables (long vs short). Syllable is an empathic way to interpret what we hear: if someone speaks not following the general patterns of the language, the hearer will feel confused and annoyed. Moreove, the kids play with the syllables, we use the idea of syllable to make rimes in poetry. We sing and we talk following certain rhythms. Without the syllable the language we speak would be impossible. However, dont ask Praat to give you the segmentation of speech in syllable. Try to hear and principally to feel the syllable. The linguistic theories simply follow what the speakers feel and think about their languages...

Klaus,

In first language acquisition, children at the stage right before they utter their first recognizable words "play" with syllables, it's called reduplicative babbling such as in bababa, bibibi. It's very rare to find babbling with just vowels or just consonants and it tends to point to a problem in hearing. Were the syllable not a reality, I think this would be a lot harder to explain. Preschoolers play with syllables later too, you will find them marching or clapping to the rhythm of the syllables. Native speakers have generally a clear idea of how many syllables there are in a word or phrase. Boundaries might be tricky sometimes (ambisyllabicity does exist) but the number of syllables is usually clear, even with syllabic consonants.

You find languages that have longer stressed syllables and very short unstressed syllables (syllable-timed English). This has nothing to do with the number of sounds but only with stress assignment in the syllable. It shows that stress is assigned to the syllable, not the vowel.

In a language like German (or French I believe), which is syllable-timed, syllable length is based on the number and length of the sounds it contains. Syllables do not differ in length because of stress but because of the number and length of the individual sounds.

In English, how long a syllable takes depends on whether it is stressed or not, not on the number of sounds it contains.

Btw, as a German speaker myself, I'm curious, how would you explain final-obstruent devoicing in German, without referring to the syllable?

For example: Rad(t) vs Rae-der? Lob(p) vs. lo-ben

Klaus,

In my view the question is a bit ill-posed. There is a more fundamental unit of sound and that is what one should focus on, to fully understand what is going on perceptually. This more basic unit is "sub-phonemic" and it is that part of the sound that allows one to discriminate each consonant (or vowel) from another. To believe this you must experience it. We have many of perceptual experiments that deeply delve into this question.

-The demos are the place to start, at:

  http://hear.beckman.illinois.edu/Files/VideoDemos

  http://hear.beckman.illinois.edu/AuditoryModels/HomePage

-And then you can read the details of how we came to all this, for example at:

  http://hear.beckman.illinois.edu/Main/Publications

-A good paper to start with might be:

  http://hear.beckman.illinois.edu/wiki/uploads/Main/Publications/LiMenonAllen10.pdf

-and good papers to end with (or read in reverse order):

Riya Singh and Jont Allen (2012); "The influence of stop consonants’ perceptual features on the Articulation Index model," J. Acoust. Soc. Am., apr v131,3051-3068

  http://hear.ai.uiuc.edu/public/SinghAllen12.pdf

Toscano, Joseph and Allen, Jont B (2014) Across and within consonant errors for isolated syllables in noise, Journal of Speech, Language, and Hearing Research, July 2014;

  http://hear.ai.uiuc.edu/public/Toscano-Allen-JSLHR-2014.pdf

Jont, it looks like you are referring to phonic features. Before linguists developed graphic systems to represent these, one historical writing system represented features within compound graphs: the hangul system, developed in the royal court of Korea in the 15th century A D.

Peter MacNeilage makes a strong case for the importance of the syllable in his book, The Origin of Speech, Oxford University Press, 2008.  He claims that it's not a coincidence that the movement of the jaws in producing syllables has the same rate of speed as chewing.  In fMRI scans, both kinds of actions light up Broadmann's area 44, which overlaps Broca's area -- the first area of the brain recognized as critical for language generation.

He goes into detail about related linguistic, psycholinguistic, neural, and evolutionary issues.  For example, the calls by monkeys and apes are limited in their variety.  But the innovation of syllables created a stream of discrete units that enable a much higher rate of information transmission.  For a tutorial on the neural and psycholinguistic issues, see the slides http://www.jfsowa.com/talks/goal2.pdf .

A syllable is a conventional unit (tool) which allows linguists to explain phonetics in the clearest possible way. However, as some other remarkable colaborators to this issue have pointed out, theory has a phenomenological distance from common practice. Let us forget not that syllabic repetition - since we are kids - is essential to learn speaking. When adults,  we have several phonetic variants due to age, dwelling at the level of the sea, in mountains, or extreme temperatures.  We sometimes speak in chunks, not syllables (or protosyllables, or postsyllables?).

Bettina and Rafael, you wrote:

"In first language acquisition, children at the stage right before they utter their first recognizable words "play" with syllables, it's called reduplicative babbling" and:

"Let us not forget that syllabic repetition - since we are kids - is essential to learn speaking."

But, what babies 'talk' might be the result of 'baby talk'.

Klaus, you wrote:

But, what babies 'talk' might be the result of 'baby talk'. 

I wrote kids, not babies, and "The three-year-old vocabulary generally contains between 300 and 1,000 words, and the child may memorize favourite songs, stories, and nursery rhymes. In rare cases, three-years-old have mastered the ability to read. Thanks for your kind attention.

Rafael, you mention "nursery rhymes". That's – in a wide sense – 'baby talk' (motherese) too.

The syllable is very useful for teaching languages, especially when teaching a syllable-timed language like English, to speakers of languages where the syllables do not change in length due to stress, like Chinese. Being able to grasp the concept of syllable timing based on stress can make a huge difference in a language learner's comprehensibility, and it helps the learners understand the pattern of English rhythm. 

Dear Colleagues,

I wonder if you have seen Halliday's study of the syllable In Mandarin - it shows the limits of trying to work with any more atomistic notion than syllable for that language.

While the roots of a debate concerning extending units in phonology go back to the prosodic theories of J,. R. Firth, Halliday focuses only on the situation in Mandarin, but with considerable detail.

My take on this and other related debates is that the rational power of atomism, important though it be in every sphere of enquiry, runs out relatively early in Linguistics.

[For Halliday on syllables in Child Language Development - see Learning How to Mean 1975 :syllabic examples in the construction of one child's semantic system]. Best, DGB

Dear Klaus

Indian languages, Dravidian or Sanskrit origin, including Malayalam, a fusion of the two, are abugidas. We have letters, called akshara or varNa, which are always written with a vowel added; in the Script the inherent letter is a (as in absolute). These are almost same as Syllable, I think; our dictionaries give the meaning of syllable as akshara. So we speak and write and you may say think syllabically. Of course this leads to problems in reading European language words.

Malayalam has an added advantage that it has chillu letters (which have been called semiconsonants/semivowels by early European Grammatists like Gundert, Pete, Frohnmeyer, etc) which can be used without any added vowel. In malayalam, we will write water as (Consonant+long a symbol)+(t with inherent a)+R (a chillu). But this means we have three aksharas and pronounce it in three syllables where as in English water is deemed to have only two syllables. Though this is a problem, the chillus makes Malayalam the language most suitable for transliteration of European languages. Also, we have orthogramph phonetic equivalent of t, lacking in other Indian languages.

Our languages have what we call sandhi (joining/junction) rules for consonant+consonant, consonanT+vowel and vowel+vowel, for making compound words. The result is also read in syllables.

All in all, we can't think of any language with out a syllable and the most useful unit, for writing and speaking

Narayanan Bhattathiri

The sound structures of language -- which include phonemes, syllables, phrases, and prosody -- support an amazingly efficient high-speed method of transmitting information in a noisy environment.  I agree with Catharine, David, and Narayanan that the syllable may be the most important of the four -- but all are important.

I also agree with Klaus that people don't pay conscious attention to syllables in fluent speech, but the same can be said for phonemes, phrases, and prosody.  In fact, people who have never learned to read can speak their native dialect with the correct grammatical distinctions, even though they have no terminology for or even conception of words, sentences, phrases, or grammar.

Even literate people who have never studied linguistics have no understanding of the structures implicit in what they read and write.  Chinese speakers, for example, often claim that all Chinese words are one syllable -- primarily because there is a one-to-one mapping between characters and syllables.  They also claim that Chinese has no grammar -- because it doesn't have the kinds of inflections that are typical of other languages. But both claims are false.

It is said that the pen (not the tongue) is mightier than the sword, highlighting the importance of the written over the spoken, though in the past when books were not available in plenty and now a days when audio is freely available the tongue is equally mighty. Any way, what I mean is that language is both written and spoken. Phoneme is the smallest contrastive unit of a language. In European language, writing may be considered as done on the level of phonemes, but we read them and speak them by combining them in terms of syllable; In Indian languages both writing and reading and speaking are in terms of syllables. This means that syllable is the practical unit of human language. Recursivity, the hall mark of human language, starts with syllables I think. This is most apparent in Languages that have 'Sandhi' rules, ie which may be essentially seen as rules for combining phonemes. In Indian languages, each such syllable may have a definite meaning too; ie. their place in combination contributes to the meaning of the morpheme, such that anywhere they occur their meaning may be present though is hidden. English too may be having this phenomenon: take the case of conditional segmentation and quasimorphemes such as fl-in  flash, flare, flicker, flame, etc. where it is linked to light, and fly, flap, flit, flutter etc to air. Fl is a syllable, and almost a morpheme; but totally neglected by Linguistics. I would say that Linguistics is yet to study syllables well, and as Klaus asked, I think this is is a regularity (or irregularity) which cannot be explained without recourse to syllables. Please do see my paper on quasimorphemes uploaded in ResearchGate

Narayanan Bhattathiri

A syllable is a phonetic and a phonological reality. So, we need that entity. Phonetically, the syllable is the necessary process of adapting the aerodynamic mechanism we use to breathe in order to be able to produce speech. The air-flow may vary in intensity and in duration (stability of the flow). In order to have some charming use of it: avoiding monotony... the process of speaking varies the characteristics of the air-flow producing units adapted to be linguistically the syllables. Languages uses that process to build up the rhythm. Phonologically, the syllable is a kind o "context", an "environment" that helps to formulate phonological rules. Therefore, the syllable is a necessary and a well adapted unit to make language more interesting.

Dear Luiz Carlos,

you wrote:

 „A syllable is a phonetic and a phonological reality.“

 Why are you so sure? Syllables – like words - cannot be seen in a spectrogram. Therefore one may ask whether they exist. Words exist because they have meaning. Syllables have no meaning.

Klaus,

Any metalanguage used to describe language of any kind depends on some humanly designed theory.  Words are just as theory bound as phonemes, morphemes, syllables, phrases, sentences, paragraphs, or prosody.  The English 'life insurance company employee' is usually classified as four words because the written form separates them by blanks.

But the German 'Lebensversicherungsgesellschaftsangestellter' is called a single word because each blank is filled with an S.  In either case, a spectrogram won't show any separation.

A morpheme is defined as the smallest unit that has a meaning.  For the German 'Angestellter', there are five morphemes, all derived from Germanic roots:  'an', 'ge', 'stell', 't', and 'er'.   But the English 'insurance' is derived from Latin roots.  That raises a question whether the Latin morphemes are still considered meaningful in English.  Since English also has the related verb 'insure', one could say that 'insurance' has at least two morphemes.

Depending on their experience, different people may subdivide the same "word" into different morphemes.  For example, the English 'hamburger' is a borrowing from German.  But many English speakers analyze it as ham-burger, and talk about turkey burgers or veggie burgers.

Whether or not that analysis is historically or linguistically "correct", it shows that even people who have never studied linguistics have a tendency to analyze words into morphemes and to recombine those morphemes in novel ways.  That suggests that they have an intuitive feeling for syllables as parts of words, many of which may have an independent meaning.

As another example, compare the spectrogram for the phrase "I don't care" as spoken by a native English speaker and a speaker with a Spanish accent.  For the English speaker, the T is represented by an abrupt end of the N sound followed by an absence of sound.   For the Spanish speaker, the sound of the N slowly diminishes up to the K sound for C.

Dear Klaus,

meaning is not really a criterion. Sounds by themselves have no meaning either. You were asking whether the syllable is more than a theoretical construct. I think there have been various comments that described examples on how syllables are a reality in language. Children acquiring language through syllables, not individual sounds, stress vs. syllable-timed languages, cultures that develop writing systems based on syllables instead of sounds, a lot of examples that I didn't quite understand. Morphemes have meaning, phrases have meaning, but we are talking about a phonological/prosodic unit here, not a meaningful unit.

Dear Klaus

As I mentioned earlier, in Indian languages, the letters (aksharas), written and spoken, are representable syllable, ie why abugidas are also called alphasyllabarics. More over, in Sanskrit and Malayalam, each Akshara ie syllable has a meaning and in some words the meaning can be derived from the meaning of the aksharas in it. This more so when they are word initial (mostly they are then treated as prefixes), but in some words they carry this meaning regardless of the position. All such has not been identified, though I think it is possible. Even in English, the combinations (called quasimorphemes; they are syllables or atleast quasi-syllables, aren't they?) such as fl-, gl-, gr-, etc does in some mysterious way contribute to the word meaning. We sa that syllables have no meaning, but the truth is that this has not been rigorously looked for. Doesn't the -syllable -asp have some commonality in the words grasp, clasp, gasp (he is actually grasping to life!), and even wasp (when a w'asp' stings, it can remains cl'asp'ed to the body since its sting has barbs; a bee sting just breaks away; so who ever gave the name could have been aware of this fact, isn't it so?). I am sure the 'namers' had reasonably good knowledge of nature and even science; only thing is that we are yet reach a level to appreciate fact. Please see my paper on Biolinguistics uploaded in RG).

Any way, from your own statement, if most syllables are proved to have their own meaning, I am sure that you have to agree that it is the most important unit.

Narayanan Bhattathiri

How a Wasp Stings

The stinger of a wasp has two parts to it, each flat and lined with barbs that stick into the skin and grab it to keep the victim from pulling free. One half of the stinger pierces the flesh, while the other pushes past it and latches into the skin. Once this second half is latched on, the other slides further in and latches on as well. The process continues till the stinger is embedded into the skin. This all happens in less than a second. The wasp then injects the venom into the victim and pulls out the stinger. Because the hooks are on the inside of the stinger instead of the outside, a wasp can pull out and sting repeatedly without tearing its stinger out. Other insects, such as bees, can only sting once before their stinger is ripped out.

Narayanan

Dear John, you wrote:

1.     „Any metalanguage used to decscribe language of any kind depends on some humanly designed theory.“

I agree. But this does not mean, that all humanly designed theories decscribe the reality (of language). Whether they do or not is what has to be reviewed again and again. Science is not possible without these reviews.

2.     „A morpheme is defined as the smallest unit that has a meaning.“

I agree to the sentence: „A morpheme is […a ] unit that has a meaning.“ But syllables are not the same as morphemes. Syllables don’t have a meaning.

Dear Bettina, you wrote:

„Sounds by themselves have no meaning.“

I disagree. Sounds by themselves have meaning. Phonemes function as meaning-discriminative units (segments). (They became segmented because of their function. They don’t exist without their function.) So – there is no other logical conclusion – their discriminative function IS their meaning (is semantics).

(Not the morpheme is the smallest unit that has a meaning, but the phoneme.)

Dear Klaus,

I certainly agree with that point:  "But syllables are not the same as morphemes. Syllables don’t have a meaning."  I never suggested or implied an equivalence between them.

The fundamental principle I have been emphasizing in several notes to this thread is that every metalanguage term used to describe any natural language is theory dependent.  That even includes the term 'word'.  For every such term, even 'word', if you try to state a precise definition that works for one language, you will find other languages for which that definition is inaccurate and requires some revision to handle the exceptions.

I'll also admit that you can state theories for which some widely used terms, such as 'syllable', can be avoided by using various kinds of paraphrase.  However, linguists use the term 'syllable' so widely because theories that use it are clearer, more general, and more succinct than theories that avoid it.

I went through the contributions again, and I still wonder whether a general linguistic concept of a „SYLLABLE“ is theoretically justifiable and whether it is practical and useful. See, what I read:

Luiz wrote:

„If you tap your back when saying a long aaaaa in continuous movement, the interruption of the air flow from the lugs will give you controlled syllables (long vs short).“

I think, it were wonderful, if it were so easy to segment syllables.

Interesting is, that an interruption might have quite another meaning  - as John pointed out:

John: „compare the spectrogram for the phrase "I don't care" as spoken by a native English speaker and a speaker with a Spanish accent.  For the English speaker, the T is represented by an abrupt end of the N sound followed by an absence of sound.“ –

Narayanan wrote, that the English word „water“ heard through Indian ears consists of three syllables: „But this means we have three aksharas and pronounce it in three syllables where as in English water is deemed to have only two syllables.“ –

John wrote, that „Chinese speakers, for example, often claim that all Chinese words are one syllable -- primarily because there is a one-to-one mapping between characters and syllables.“ –

Narayanan wrote: „In European language, writing may be considered as done on the level of phonemes, but we read them and speak them by combining them in terms of syllable; In Indian languages both writing and reading and speaking are in terms of syllables. This means that syllable is the practical unit of human language. […]

In Indian languages, each such syllable may have a definite meaning too […]“ –

Again Narayanan: „…as fl- in  flash, flare, flicker, flame, etc. […] is linked to light, and fly, flap, flit, flutter etc to air. Fl is a syllable, and almost a morpheme; but totally neglected by Linguistics.“ –

Luiz: „Phonologically, the syllable is a kind of "context", an "environment" that helps to formulate phonological rules. Therefore, the syllable is a necessary and a well adapted unit“ (What do the phonological rules help? Whom??)

Dear Klaus

By now, all would have understood my viewpoint: syllables are very important, and the backbone of Indian Languages, even though our Linguists can speak of phonemes, etc, quite wisely. Our alphabets, whether Dravidian or Sanskrit originated, is invariable taught in syllables (ie vowels or consonant+vowel, and the default vowel is a (as in adamant :).

Regarding use other than in our context, many practical applications and theoris seems to have been built around the syllable. For example

1) A speech recognizing device performing speech syllable recognition and language word identification has been patented (http://www.google.com/patents/US5675705)

2.) Syllables are as distinctive as phonemes in spectrograms (http://www.cslu.ogi.edu/tutordemos/SpectrogramReading/spectrogram.html)

3.)Syllables can be classified according to Acoustic-Auditory features and may be better for Automatic Speech recognition than other linguistics units (http://www.cstr.ed.ac.uk/downloads/publications/2003/wester.2003.1.pdf)

4.) Word classification according to syllables may be more useful for computer scientists in some areas (https://hal.archives-ouvertes.fr/lirmm-00671499/document)

The only thing that syllables lack is definite link with meaning. It is not their fault; the real fault is that no one bothered to do it. Also, how do you find the meaning of the syllable 'gh' in a word like 'through' where it is silent? It is not the fault of the syllable that it is not pronounced.

Actually syllables can be classified into three

1. Those with a well define meaning such as a, ab, etc.

2. Those with ill understood meaning (quasimorphemes) such as fl-, gl, etc., that I mentioned earlier

3. Those the meaning of which have never been looked for.

If the third is researched well and their meanings establshed, they will become Morphemes, isn't it.

Narayanan Bhattathiri

Aren't there some dialects of Japanese in which accent placement depends on syllable counting (rather than mora counting)?

Syllable seems to be more unique to some languages than others. In tone languages such as Yoruba, tones are borne by the syllables. In terms of meaning, we should know that a number of factors are responsible. Just as has been noted by other contributors, syllable may not be associated with meaning in some languages, but it should be noted that syllable is an autosegment. In Autosegmental Phonology, a non-segment item like syllable or nasalisation can affect meaning just like any other phoneme. We just have to know that many things that a not possible in English other widely spoken European languages are very common in some other languages. Linguists are more interested in a linguistic feature than in its occurrence in popular languages.

My website has a different URL (see my resp near the top of this page)

https://auditorymodels.org/

The demos are now at:

https://auditorymodels.org/index.php?n=AuditoryModels.HomePage