Phonemes are three dimensional: each phoneme has a specific 1) manner of articulation, 2) place of articulation, and 3) voicing...

  1. 1)MANNER is all about what kind of friction or resistance the up-coming airstream is caused to meet by closure or partial closure or lack of closure in the vocal tract - e.g, full closure with p, partial with f, and none with h. Note: There could be a complete closure in the mouth, but with a re-direction of the airstream through the nose - e.g, n. This is called nasalization - as opposed to oral, e.g, d.

  1. 2)PLACING is all about where the closure or partial closure is located - the most prominent of English being the following: with both lips (bilabial) with b, p, m, w;  with the bottom lip and top row of teeth (labio-dental) with v, f; with the tongue in between the teeth (interdental) with th*; with the more or less tip of the tongue touching the ridge behind the top row of teeth (the alveolar ridge) with d, t, n, s, z; at the hard palate (= the middle of the “dome” of the mouth) with consonantal y (i.e, as in “yuck” but not “lady”); between the alveolar ridge and hard palate with sh; a movement from an articulation with the alveolar ridge to the hard palate with ch and j; the velum (a.k.a. soft palate) with k, g (as in “get” but not “gem”); a free flow of air right out of the throat (specifically, right out of the glottis) with h and all vowels; and the tongue is in various special shapes with the “liquids” l and r (the former basically alveolar, the latter a long story LOL - no wonder it’s often omitted in many dialects and languages!). There are also placings that are co-articulations. One we talked about - ch and j with movements from the alveolar ridge to the hard palate - are affricates... which, basically, are two sounds put together, possibly two sounds which in the given language may each also be independent phonemes, but by being pronounced so closely together are then a whole other phoneme. But there are other forms of co-articulation: e.g, taking a velar consonant like k and rounding the lips (labial) when pronouncing it: qu. This would be called labio-velar. Note how qu is basically the sound k plus the sound w. One can conceive of k plus y as another good example of co-articulation (it would be close to the ch-sound).

  1. 3)VOICING is all about the state of molecules in the airstream - whether the vocal cords are causing them to vibrate or not: that is the basic distinction here. This is a feature written [±Voiced] in the binary “code” so favored by modern linguists. That deep, deep area of the throat - the glottis (where h comes from) - is where the vocal cords are located; and it is when they vibrate that the airstream vibrates, [+Voiced], and when they do not vibrate that the airstream is not in vibration, [-Voiced]. The area of the glottis can cause other disturbances in flow of air - such as by stopping the air flow  short with a sharp, sudden sort of “pinch” of the glottis (glottalization), or a relaxed “lazy” extra letting go of the stream of air even after the closure has been released (aspiration). (NOTE: The vocal cords are not physically capable of vibrating with glottalization, though may do so with aspiration, but usually do not: “voiced aspiration” is also known as “murmur” .)


    On the hierarchy of phonology - as we are considering it from the ground up - the next consideration is what phoneme in this or that given language can or cannot be positioned together: this is known as phonotactics. For instance, in English a word beginning with mg would be a violation of the phonotactics; but in Russian this would not be (e.g, mgora, ‘fog’). On the other hand, in English we could have a word beginning with sp, but in Spanish we could not - and to fix that problem, Spanish would add e before the sp: compare English Spain with Spanish España, basically the same word, one pronounced English-style, the other Spanish-style. This “style” is phonotactics.

    Languages adapt words in accordance to their respective phonotactics. Adding a phoneme to solve a problem sound-cluster is epenthesis (Spanish Spaña > España). Alternatively, there can be deletion: e.g, Greek psyche is pronounced without the p in English. And there is also mutation: Tlingit in English is pronounced (though not spelt) Klingit. To illustrate further: Japanese phonotactics disallow almost all consonant clusters - hence my first name, Francis, would be adapted by Japanese grammar to become Furansisu (NOTE: ns is allowed).


    Notice how the Japanified version of my name has four syllables (, whereas the English version has two (Fran.cis). Indeed, phonotactics influence syllabification, i.e, how the phonemes come together in syllables. For instance, let’s say we have two consonants together, pt... One of them might be placed at the end of the first syllable, and the other at the beginning of the second syllable... rip.tide would be acceptable syllabification; but ript.ide or ri.ptide would not be. ri.ptide would not be for the same reason it would not be acceptable in English to have a word that starts with pt. ript.ide would not be acceptable for a different reason, which is as follows...

...Generally speaking, in the languages of the world - i.e, in Language per se - the ideal syllable is one beginning in a consonant and ending in a vowel. This is why ript.ide is not acceptable whereas rip.tide is. In other words, why have syllable “ide” when you could have “tide” (beginning with a consonant)? Japanese is very faithful to such syllabification, as we have seen with - which is more faithful than Fran.cis. A CV syllable - i.e, a syllable of one single Consonant and one single Vowel - is a “core syllable” .


    We have now moved up the phonological ladder from...








However, just as rudimentary as features are to phonemes - and, indeed, one or so features of its own; - and an element of syllables without which there could be no syllables - is the MORA. The mora is a unit of time, like a single one-beat note in a musical score. The feature [±Syllabic] means whether or not a phoneme can be at the heart - nucleus - of a syllable: vowels are this, and at times so are certain other consonants, like the “liquids” l and r, or nasals m, n, etc. The feature [±Long] represents whether or not a phoneme is short (i.e, lasts for one mora) or long (i.e, lasts for more than mora). The long vowels in Classical Latin - written ā, ē, ī, ō, ū - would be, phonemically, the same vowels as what in Latin are written a, e, i, o, u (the short vowels) - except that there is ONE (and only one) distinctive feature differentiating the two sets: the feature [±Long] is plus for the former and minus for the latter.

    How many morae (a.k.a. moras) a syllable has directly effects - indeed, determines - whether or not a syllable is short or long. The core syllable is short by definition, CV; but to make it long, we could write it as CVV - which can mean there are two short vowels (and thus two morae), or that there is one long vowel metrically equivalent to two V’s. Another way to add a mora to a syllable is to have it end with one or more consonants - it makes no difference how many, one or a few are equal to one mora. Thus, CVC has two morae, as does CVV; CVVC has three morae, but CVCC just two.


    Metrically speaking, the first entry of the phonological ladder is the...


...then comes the


...and then the


...then the

phonological word (a.k.a. p-word)

...then the


...then the


and then the

intonation of discourse .

    Right now, I would like us to consider only the first three rungs of the metrical ladder: mora > syllable > foot. This is because, in my opinion, the languages of the world can be typologically classified in accordance to which of these “rungs” a given language’s grammar regards as the fundamental distinctive metrical unit. Which metrical unit a grammar “says” should be primarily heeded can have a great affect on the overall accent of that language.

  1. 1)MORA-BASED. When listening to Japanese, one hears an accent of mora-by-mora pronunciation - and Japanese poetry seems to reflect this, I believe (just listen to the Japanese version of the Prajna Paramita Heart Sutra, and compare it with the original Sanskrit).

  1. 2)SYLLABLE-BASED. When one listens to Romance languages, one hears an accent of a syllable-by-syllable pronunciation - there is more of an accent based on changing the pitch of different syllables instead of a stress-type of accent that may  compromise the care with which each syllable is pronounced, i.e, in terms of length, volume or even faithfulness to the original quality of phonemes (the distinctive features may not be as faithfully expressed or adhered to).

  1. 3)STRESS-BASED. In Germanic languages, we have such a stress accent. Here, the foot is the most important metrical unit. The Germanic foot is formed from either one long stressed syllable or a stressed short one followed by another syllable of whatever size. [See The Germanic Foot: Metrical Coherence in Old English by Aditi Lahiri and B Elan Dresher; pp. 251-287 of Linguistic Inquiry, Volume 22, Number 2, Spring 1991, MIT.] The second syllable is pronounced less distinctly, and in the passage of generations might be deleted. Often, if there is a consonant located between the two syllables, it becomes just as much part of the first syllable as the second, i.e, it is pronounced in such a rapid way that the closure of this consonant is the end of the first syllable, and the release of that closure is the beginning of the next syllable: the “t” in “atom” is a good example. “t” here is ambisyllablic, i.e, belonging to two syllables. Thus, in Germanic languages the syllable boundaries are not always clear, and the syllable itself is not always thought of as worth preserving. What is key is that a certain foot structure is adhered to.

    Romance is best for rhyming poetry where the amount of syllables per each line of verse are to be of a certain number; and the end syllable of the rhyming words are clearly enunciated. -Whereas Germanic is best for alliterative verse where the first syllable is what is most important, and is stressed, often to the extent that the following syllables fade and fade until the word is over. (And the Japanese haiku counts morae - not syllables!) It seems the Germanic accent makes word-boundaries more distinct than they would be in Romance.


    When we talk about words, we could define words semantically - this is what the dictionary is supposed to do for us; or we can define words phonologically, which, basically, is how all the stuff about accent we have been talking about gives a certain cohesion to groups of phonemes to put them into a metrical unit that may or may not coincide with the semantic entity of the word: this  is the p-word (i.e, phonological word). Sometimes, two semantic words may come together phonologically to be pronounced as one p-word - and, in time, one of these semantic words might end out being, say, a suffix of the other word. This is called grammaticalization.

    Words in any given language may have more or less of an independent phonological identity. In the Celtic languages, for instance, there are sound mutations on the edges of words depending on which words those words occur with. Thus, the phonemes at the beginning or end of words may change with each phrase!


    -But now we have risen to the phrase level. One may think of special accents for noun phrases as opposed to, say, prepositional phrases - i.e, intonations tied to parts of speech at the phrasal level. Higher up the ladder are clauses and sentences. We know from English how different types of clauses or sentences can carry different intonations - e.g, the rising intonations of questions and the intonations that “!” can stand for. On the very top of the metrical ladder - and, indeed, the phonological ladder - we have the intonations that have functions or meanings in the context of a discourse: e.g, an angry tone, a polite tone, epistemic tones showing doubt or confidence - so on...


* There are actually two types of th: one is voiceless as in “thin” and “through” , the other voiced as in “then” and “though” .

Contents of this lecture:

  1. 1)Introduction

  1. 2)Phonology (where you are)

  1. 3)Lexemes

  1. 4)Morphology

  1. 5)Syntax

  1. 6)Conclusion

Click here for a glossary for an artfically constructed lexicon (^_^).