Syntax - “with Tact” {work in progress}


    Let us wrap up the discussion on the components of grammar by mentioning syntax. (We will not get into pragmatics, because that would be entering a gray area that is less linguistic and somewhat psychological and sociological.) Basically, syntax is the order of elements in a phrase, clause* and sentence, and these elements are words, more or less (sometimes they may be particles and grammemes, but we won’t open the can of worms constituted by the question of whether these are words too). It is logical to keep the predicate together - i.e, the object and the verb. Yet, that is not always what happens in Language. So, linguists - perhaps mistaking Language for Logic (just like the prescriptive grammarians do) - say that those languages which split the predicate have, underlyingly, predicates that are not split. They say that before a sentence is uttered, it goes through levels of transformation, from having a non-split predicate, to having a split one. Most notable of the languages that split the predicate are the Celtic and Semitic languages. All of them are known for having the verb first, then the subject. However, many languages - like the classic Indo-European languages (Latin, Greek, Sanskrit, etc.) have a suffix on the verb that stands for the subject (like a pronoun would); and so it might even seem that Language has a universal tendency to keep the subject right behind the verb. Maybe. The point I am trying to make here is that Language is its own thing, a separate species from Logic, just like Math. But what I am ultimately getting at here is that Language stems from the Intellect/“grasper,” and not from Logic. Perhaps Language and Math and Logic are all siblings. Or perhaps Logic is the offspring of Language. But Language is not the offspring of Logic.

    We can organize languages into distinct linguistic types by means of syntax. Coming from the same school of though as Schleicher with his threefold classificatory system based on morphological types (monosyllabic/isolating; agglutinative; and fusional) Wilhelm von Humboldt* nots a certin other threefold distinction - one which fits the polysynthetic/incoporating type into the classificatory system. This way of classifying languages - this morpho|syntactic way - takes into account how the morphology reflects the grammar of the whole sentence. Here we have the threefold distinction of 1) monosyllabic/isolating which do not express syntactic relations morphologically, but purely syntactically (syntax used to express syntax - go figure! LOL), e.g, through how the words are ordered or positioned, or through special words which - although distinct lexemes, are used to denote grammatical/syntactic categories; then there is 2) langugaes which use morphology to express grammatical relations, such as 2a) agglutinative languages which express such relations but - as Schleicher pointed out - at the expense of word-unity, and 2b) fusional languages which express such relations but without - as Schleicher pointed out - loosing word-unity. Schleicher talked much about word-unity (Worteinheit, literally ‘word-oneness’), but Humboldt brings up the idea of sentence-unity: all three have sentence unity to one extent or another, but (1) is the least unified because its sentence is divided into the most lexemes; (2) is more unified because it uses less lexemes; then (3), which is the most unified, and in its ideal is so unified that word-unity is one and the same as sentence-unity - one lexeme = one sentence (with morphemes, though). Thus, we now have the necessary missing parameter to update Schleicher’s system to now be fourfold... Before, we just had the considerations of A) morphological expression of grammatical relation, on the one hand; and B) preserving word-unity, on the the other hand. Now we can add C) achieving sentence-unity... This is the sole domain of polysynthetic/incorporating languages (in their ideal form)...

                            morph-    word-    sentence-

                            -ology       unity       unity


isolating                 no            yes           no

agglutinative         yes            no            no

fusional                 yes            yes           no

polysynthetic/      yes            yes           yes


Interestingly, in Schleicher’s classification, he compares his three classes of languages to what he would consider developmental stages in some sort of evolution towards perfect (which is thinks of as represented by the fusional class), and in this regard he compares the monosyllabic/isolating class to the non-animate mineral kingdom; then the agglutinating class to the semi-animate plant kingdom; and finally the fusional class with the fully animate animal kingdom. In the spirit of these analogies, what should we compare the polysynthetic/incorporating class to? How about this: super-conscious god!!! LOL But seriously, that would be logical...

    Of course, Schleicher probably wanted, and felt he had to, follow the evolutionary cycle as described by the Hegelian Dialectic: thesis, antitheses, synthesis. In this way of thinking about things, history is supposed to follow this dialectic. One thing happens, then its opposite happens, then something happens which encompasses the better parts of both: monosyllabic/isolating = thesis; agglutinative  = antithesis; and fusional = synthesis. But by adding polsynthetic/incorporating to our classificatory system, we seem to destroy its compatibility with the philosophy of Hegel...


...we make the fusional class, which is the synthesis of the previous two classes, be, in turn, the thesis for a new level in the dialectic progression. After fusional as a thesis - which would be

                         morph-    word-    sentence-

                          -ology       unity       unity

fusional                yes         yes           no

we then have

                         morph-    word-    sentence-

                          -ology       unity       unity

???                       no           no           yes

??? would be the antithesis of fusional. As the synthesis of ??? and fusional, we would at last come to polysynthetic/incorporating...


isolating             → agglutinative



                         fusional →    ???





    In any case, we know at least that the classification of languages - and thus Language in itself - is not just about morphology (or the lack thereof). Syntax is real, syntax must be considered, dealt with. Indeed, one around could argue there is no such thing as morphology, that in reality the nuts and bolts of Language break down to phonology, on the one hand, and syntax, on the other; and what appears as “morphology” is really a bunch of phonological processes acting upon lexemes, making some lexemes more clearly articulated, or less clearly articulated (the latter would be morphology’s morphemes). And yet, it is precisely this difference in the faithfulness which the phonology “decides” to articulate some forms as opposed to others that suggests there is indeed something else going on, namely, morphology. And if on this basis we posit that there is such a thing as morphology, then we can account for another phenomenon: that is, in a given language, one may find two systems of ordering elements: one on the word level (such as amongst morphemes attached to a lexeme), and one between lexemes, on the sentence-level, i.e, syntax and word-order. The two in any given language do not necessarily order constituents in the same order.  That is how we know there are these two systems. And in this way, the very existence of syntax is confirmed by the existence of morphology. In other words, this theoretical framework fits, though not necessarily ultimately unproven.

    The order of elements, and the way the particular patterns of order may be similar or dissimilar between morphology and syntax, is itself a parameter in the classification of languages. The basic possibilities basic possibilities for syntactic order are as follows - and, indeed, a language can be classified according to which of these orders it has...

SVO = Subject, Verb, Object (e.g, English)

SOV = Subject, Object, Verb (e.g, Latin)

VSO = Verb, Object, Subject (e.g, the Celtic and Semitic language families)

...and here are the others, which we will not concern ourselves with, since they are either non-occuring or rare...


    These word-orders can be divided into two constituents: the V and the O form the Predicate; and S is the Subject. In both morphology and syntax, when a number of elements are put together, we must consider which one is the main one, i.e, the governing head - and in the case of VO or OV, that head is V. And so, a language that has VO has “head-governed” order, and one with OV has “governed-head” order. But when one compares this syntactic order to that of morphology, one may find “governed-head” morphology but “head-governed” syntax, or vice versa...

EXAMPLE: from English...

(I) walked the dog.

walked = Verb (governing)

the dog = Object (governed)

Syntax = (S)VO,



walked = walk + ed

walk = governed

(=lexical root of word,

Scheicher’s Bedeutungslaut)

ed = governing head

(=grammatical morpheme,

Schleicher’s Beziehungslaut)

Morphology = governed-governing.

    This itself confirms the existence of syntax and morphology as separate and distinct entities. Yet this is not to say, however, that a different a different component of Grammar, that is, phonology, cannot play a role in connecting both: for example, the phonology of a given language might “prefer” to always stress the lexical root per se. By ‘lexical root’ can mean that number one morpheme of a word, the one that carries its lexical meaning, i.e, “walk” in “walked;” and the lexical root of the sentence, as we would have to conclude from the phenomena of polysynthetic/incorporating languages, is the verb: this would make “walk” not only the lexical root of the verb “walk + ed,” but the lexical root of the entire sentence.

    One can account for the shift in word order in Germanic (verb second) as it came out of Proto-Indo-European (verb last) as parallel and a direct result of Germanic’s shift of accent as it developed out of Proto-Indo-European from a free and floating pitch accent to one of a fixed stress accent on the first element (trochaic) and on the lexical root. Now, it is an issue in the study of the history of Germanic whether this accent was morphologically based - i.e, the target of stress being the lexical root - or phonologically based - i.e, the target of the stress being the initial part of the lexeme, irrespective of whether or not that initial part was the lexical root. I say there were both of those tendencies/preferences in Germanic, and ideally, the grammar would want both to be satisfied - and, indeed, sometimes the order of constituents would be altered to cause both constraints to be satisfied. This, I argue, is what happened in Germanic syntax. The Proto-Indo-European word order of SOV, where the lexical root is last, was first changed (in the minds of the speakers before physical utterance) to VSO, where the root is first to satisfy both preferences. [These “preferences” are what Optimality Theory calls “constraints” - see the works of John McCarthy and Alan Prince.] Although this would divide the predicate, that in itself is not necessarily an object because divided predicates do occur in real languages of the real word, e.g, those of the Celtic and Semitic language families are VSO. Still, of course, it must be a universal preference of Language in itself to keep the predicate undivided. This preference obviously does not take on a  particularly high priority in the Celtic and Semitic languages, but in Germanic it did: enough so that the grammar solved the undesirable split in the predicate of VSO with Germanic’s famous verb-second (V2) phenomenon, i.e, the verb always coming in the second place of a paratactic (main) clause by means of O or S being moved (just before the clause becomes uttered) to be in front of V...


                                ↙       ↘

                        SVO               OVS


    Since OVS is either rare or non-occuring int he world’s languages, it is fair to say it is more ‘marked’ (=less desirable) than SVO. In other words, the grammar would have to have a good reason to “be convinced” to use use OVS instead of SVO. A language would “want” to follow the preference of keeping the subject initial. However, in a given context, a language might prioritize stressing a certain element in a discourse (this would be what I would call a “pragmatic constraint,” if I were trying to keep to a more traditional Optimality Theory framework), and in this case a need for emphasis on, say, the object may override the preference to keep the subject initial. Hence, in Germanic - and German - we may also see OVS.

    In Germanic, subordinate/hypotactic clauses retain the older SOV order. I argue that this is because the hypotactic clause itself is analogous to the weak part of the Germanic  foot. (This entails it would be rare for a hypotactic clause in German/Germanic to precede the paratactic one, for that would be an iambic stucture, rather than that of the Germanic foot, trochaic.)

    Incidentally, the difference between a paratactic/main/coordinating clause and a hypotactic/subordinate one, is this: “John went to see the Dutchess” is paractic. Then, if I follow that clause by “because he had to feast his eyes,” then I would have added a hypotactic clause - which is to say a clause which by itself cannot be a sentence, and which depends on, is governed by, the main or paratactic clause.

     Now, when there is a language that at any given utterance may have the object first or the subject first, how is the hearer supposed to know whether s/he hears an object or a subject? This gets us into another issue, another one concerning the interaction of morphology and syntax. In fusional and agglutinating languages, the nouns tend to be marked, morphologically, for their syntactic roles in the sentence. These markings are called cases. Examples (from Latin):

dominus (lord) = nominative singular (the nominative case is generally that which is used for the subject - although there is an another type of language, the ergative-absolutive type, which we shall discuss shortly).

dominum (lord) = accustaive singular (the accusative case is generally that which is used as the direct subject - in a nominative-accusative language, ass opposed to an ergative absolutive language)

There are many other cases in Latin, but we do not need to concern ourselves with them here. We simply say that in addition to the TYPE four morphological types we have discussed, the three that are capable of being case language (fusional or agglutinating or polysynthetic/incorporating) can be divided within themselves between whether they designate subject vs. object with the nominative-accusative system, or with the ergative-absolutive case system. We can describe what the difference between the two types is as follows...

nominative-accusative system...

nominative = subject of the verb, i.e, in the sentence “the elephant eats beans,” the elephant is the agent doing the eating, and the verb reflects this by being conjugated for the singular, to match how the noun “elephant” is singular; it is irrelevant to the verb that “beans” is plural.

Whether or not a verb is transitive, e.g, “eats [beans]” or intransitive, e.g, “walks”, would not change the case: “the elephant” in “the elephant eats beans” is just as much nominative as in the sentence “the elephant walks.” But in an ergative-absolutive language, this would not be the case...

ergative-absolutive case

“The elephant walks.”

the elephant=absolutive

“The elephant eats beans.”

the elephant=ergative

the cans [of soda]=absolutive

Thus, what we call the subject in “the elephant walks” does not receive the same case-marking in an ergative-absolutive language as what we call the subject in “the elephant eats beans” - indeed, what we call the direct object in “the elephant eats beans” receives the same case-marking as what we call the subject in “the elephant walks.”

To understand what is going on in ergative-absolutive languages, think of it this way: you do not talk in these languages with active forms of the verb like “walks” and “eats,” but with passive forms like “is walked” and “is/are eaten.” Therefore, in an ergative absolutive language you do not literally say “the elephant walks,” but “the elephant is walked;” nor do you literally say “the elephant eats the beans,” but “the beans are  eaten by the elephant”.

Of course, in our examples we just opened a whole new can of worms: namely, the literal versions of the ergative-absolutive refer to an action that has already taken place - not necessarily the past tense: it is really the perfect aspect. This shows how the verb system connects to the case system.


Contents of this lecture:

  1. 1)Introduction

  1. 2)Phonology

3) Lexemes

4) Morphology

5) Syntax (you are here)

  1. 6)Conclusion

* For some basic types of clauses, see my Atlantian Syntax - click here.

* pg. 128 of On Language: The Diversity of Human Language-Structure and its Influence on the Mental Development of Mankind; translated by Peter Heath;  1989, Cambridge University Press, Great Britain.





The Story of “It”

“It seems to me...” one might say; or “it is raining,” is another example of this mysterious It. What is It? A syntactician would answer: “it” is nothing but a dummy subject, i.e, a word used to fill a syntactic slot just for the sake of filling it. In other words (according to this point of view - a dry non-mystical and all-too-perfectly rational one) there is no real subject to such sentences, just a technical one to satisfy the “red tape,” as it were, of theories of syntax...

So let us investigate what is going on here. Let us do this by gradually gathering our bearings...........

  1. (1)What is the SUBJECT?

According to Bernard Comrie (in his excellent and supremely eloquent book Linguistic Typology and Language Universals) the subject of a sentence can be defined - ideally, at least - as where the semantic categories of Agent and Topic intersect. In some languages, like in Japanese, the topic itself is marked off by a case marker, “wa”, and the agent with “ga” - and a sentence may have an agent or a topic or both or none. But in English, the categories of topic and agent are con/fused, and it is this fused form that the word Subject refers to (i.e, Subject is a semantic domain that encircles the smaller domains of Agent and Topic). And so in English we have this entity Subject at the beginning of the sentence or clause, generally speaking (I will try to keep things simple for the sake of ever ending this discussion)...

  1. (2)But what if there is no AGENT?

“What is this It that rains?” one might ask. Is there some mystical entity - too big for us to grasp - that is being referred to? Such as some great “it” that may be the flow of a musical score that one flows along with, feeling or even knowing what the next notes will be, even if one has not heard the score before, because one is “with it” and “flowing with it”? “Surely,” one might ask, “Everything has a cause...” And that which causes or is the cause is a fine way of defining Agent.

  1. (3)Is this It supposed to be a TOPIC?

But what is the topic - I mean, the whole point of mentioning what the topic is, is to inform people of what is being talked about, what is in the spotlight of the sentence, as it were - i.e, for example, “in regards to...” is a good way of translating Japanese “wa”... But what is this It supposed to do? So we are brought again to ask what it would be as an agent...

(4) A Case for “Dummy Subject”

In archaic English, there was the expression “methinks,” and “think(s)” here was actually an older word for “seem(s)”. And so, in this phrase, the subject is left out. Is not this phrase of English a vestige from an ancient phase of English (a.k.a. Anglo-Saxon) when its grammar did not call for all its syntactic slots to be filled by spoken words? For, once upon a time, in Old English (=Anglo-Saxon), the language was more like Latin or, say, Polish, where, for instance, one habitually leaves out pronouns where the person/number is clear. We do not do that in modern English, even where, for instance, with the word “am” it is completely clear that the person is 1st and the number singular - i.e, the person and number the word “I” stands for. We redundantly add “I”. Why? Just to fill a syntactic slot! That is of high priority in Modern English. We even mark off the beginning of noun phrases - just to mark they are phrases - with articles (the, a(n) ) where other languages, for instances the aforementioned Latin and Polish, do not even possess such words... And is that what this “it” in “it seems to me...” and “it is raining...” means?

  1. (5)So-called “dummy subjects” may have lexical rather than syntactic significance.

We also have another “dummy subject” in English: “there.” Perhaps this “there” is a true dummy subject. We use it in expressions like: “there is a...” “there” really goes with the verb “is” (or “are”, “was” so on) - it goes with “to be” just like the prefix “ad-” went with Latin’s forms of “be” (e.g, “est”, “sunt”) in order to show this was a special form of being - “for there to be” is an infinitive phrasal lexeme in the lexicon of English, just like the single-worded - but prefixed! - “adesse” of Latin would be in its lexicon.

  1. (6)“It” as not  a dummy subject but an alluding to the Ungraspable...

The German form of “there is...” is “es gibt” - which means “it gives...” What gives? Are we again in the situation where a dummy subject is used to fill a semantic slot? Or is it the case that no such dummy subject was used in any of these examples?  If we think again how in archaic English one could say “methinks” and leave out the “it,” is that really simply because “it” would have been a mere filler of a semantic slot? If there really were an “it” that was an agent - in languages like Old English and Latin and Polish - it still would only be stated for marked emphasis. The agent can exist and be unstated. English grammar, on the other hand, seems to be an organism that does not “trust” that enough information could be expressed by a single verb that a pronoun to represent the subject could possibly not be in order. So, even though such forms as “am” are clear as to what the number and person of the subject is, many other forms of verbs are not - most, in fact - so it is easy to see how a generalization developed in the language to always use a pronoun if there is no noun already indicating the subject.

In other words, the syntactic argument of “dummy subject” can be wrong, conceivably. Something may be going on that is hard for our intellects to grasp - indeed, impossible, because we are talking about something from which our intellects spring - from Nature or God or Goddess or the Universe or whatever kind of label one tries to strangle “IT” with (Zen often uses just “it”).......... There is something greater than our intellects.... encircling our intellects; a greater domain of meaning.......

After all, why in the German form of “there is...” - literally “it gives” - is such an agent-implying word as “gives” used? Everyone knows that giving takes a giver!

  1. (7)Possible Conclusion...

There are no dummy subjects!!!

(8) Philosophical/Semantic Ramifications

Here we are touching upon the philosophy of Idealism - as opposed to Particularism. But in this kind of Idealism, this “IT” would be the sole Ideal, i.e, the biggest of semantic domains, the one that encircles all the others - which would be but fractals of the one and only true semantic domain. This ultimate semantic domain could be called “the ground of being” or consciousness (in the sense of awareness, not necessarily cognition or anything that fits into consciousness, that is encompassed or presupposed by it)... Put in a Hindu sense, one could refer to it as Brahma or Brahman or Atman... In a Zen sense, Mind or just It... The Tao of Taoism... God in the unutterable sense of the ancient Hebrews, i.e, that Ultimate whose names could not be spoken (that is why the nonsense word(???) YHVH is used, and the voice from the burning bush told Moses just to tell the people that “I am,” sent you - i.e, the ground of being; and it is the true meaning of “do not use the Lord’s Name in vain)......... In other words: Do not bother to grasp the ungraspable, to encapsulate in words what does not fit inside them... Such is VANITY!

  Nominalism/Particularism is true up until the point of the category of categories - let us label it Tao.

“It is raining.” = “Tao is raining.”

Plato thought the Ideals were various and the Mind could remember them... But he was wrong: Mind itself was the Ideal -----> and the ONLY one!!!!!

It is a dimension of the Universe.





In the section on phonology, we talked about phonological hierarchies - two kinds in fact, one segmantal, and one metrical. Here, in regards to syntax, we can make such a hierachy as well. It will be just as much phonological as it would be syntactic - for, as we have seen, phonology permeates every aspect of grammar... after all, language by definition is spoken.

The first/lowest (i.e, least category-encompassing) rung of the ladder is Clitics. Clitics are little doo-dads that lean either forwards or backwards onto an actual word (lexeme). The word “not” takes on the clitic form of orthographic “n’t” when it leans back onto “do”. When clitics lean backwards, then are ENclitics. Ones that lean forwards are PROclitics. This rung is indeed a grey area that overlaps with morphology, in addition to being sopping wet in phonological alterations (words become suffix-like or prefix-like). These clitics are not to be confused with affixes, for they are capable of attaching not only to a lexeme but to a whole phrase, as well - it is the phonology here that determines what they do r do not attach to; indeed, the whole metrical SWEEP of an utterance may gather up a certain governed words and and cause it to seem to lean in or lean forwards onto the governed particle. It is more that they are subsumed, and PULLED, rather than they are leaning; for they are unstressed, as a rule - otherwise, there would be no worn-away quality in their pronunciation, i.e, there would not be full forms lexically. However, granted, the lexical form can disappear in time, as was the case with the apostrophe S in English. It attaches to a whole noun phrase, cf. “Jack and Jill’s” as opposed to the genitive S of related language German, “Jacks und Jills,” where S must attach to each individual noun. The German S is a suffix and, as such, a matter definitely for just the mophology - as opposed to the clitic which is a tad syntactic because it takes the whole phrase into account. 

One might think the next rung of the ladder would be words (lexemes), but no, not so fast... First we have little PARTICLES that we may throw around the sentence, or around words - they are like affixes in agglutinative languages with a transparent morphology, but instead of being attached to a lexeme, they are attached to a clause, and may even refer to the entire universe of discourse directly, as in the case of modal particles or, especially, discourse markers. These particles can resemble clitics, as well, in that they many be attached to a phrase rather than a lexeme, as in Japanese particles.

The difference between clitics and particles is largely a phonological one, though - in a way analogous to a fusional vs. an agglutinative language, respectively. But clitics really do have to be on the lowest rung, and are not tied with particles, because with clitics we see a prime syntactic constituent, i.e, a word/lexeme, being broken down, through prosody, to start to appraoch being a morphological entity, i.e, an affix. (Affixes, then, we could list as even lower on this hierarchy than clitics; but then we would no longer be talking about syntax, and would have backtracked to our previous discussion of morphology.)

If there are any two rungs of the latter that are tied and can be conflated, it would be particles with the next rung - lexemes. Are lexemes really different? Well, particles are functional referents, grammatical or pragmatic, whereas lexemes speak of things or actions, to put it simplistically. And so - whereas the difference between clitics and particles was based on phonology - the difference between particles and lexemes is based on the semantics.

After the lexical rung, we come to the phrasal - which you must recognise from the section on phonology. We all know of preposition phrases, and let us keep in mind noun phrases, and so on - basically, just think of a part of speech, and that part of speech can be made into a phrase...

noun make noun phrases; verbs make verb phrases; adjectives make adjective phrases; proposiitions make prepositions phrases...

Yet the plot thickens: adjectives are often constituents of noun phrases; and noun phrases of verb phrases; and so on with such interwovenness until you come to the CLAUSE - which itself is a sort of phrase when the sentence has more than one clause. The senential clauses are combined in meaningful ways in the discourse, such as when you write an essay and the first sentential clauses constitute an introductory paragraph; and then, say you have about three paragraphs of which each issultrate your point; and then a concluding paragraph... all of these sentences came together into paragraphs which came together into one universe of discourse (albeit, of course, one in which you were talking to yourself on paper, hahaha). In real life, i.e, in real language - guess what? The universe of dicourse is jut as much connected as in an essay, but in more of a complicated way, because it involves to people (at least), and it is always a messy first draft. Fortunately we are getting out of range of syntax here and I can just finish ths damned paragraph. hahaha.

The point is that clauses come after phrases, and sentences (sentential clauses, as I call them) after that - and then the entire utterance. Phonology is present all the way, from “underneath” the syntax (distinctive features, segments, metrical units, morphology’s affixes, whathave you), and BEYOD the syntax - into the realm of pragamtics and discourse, where phonology  in the form of intonations plays a great role

ALGHOUGH, not necessarily so much in all languages as much as in English. In German, for instance, there are many modal particles that play the role of our Engilsh/American intonations....

Conversely, the intonations we English-speakers think of as reserved for syntactic and, especially, pragmatic nuances (such as in questions vs. statements in the former, and things like doubt or certaintly, whatnot, in the latter), other languages, like e.g. Chinese, may make use of these TONEMES as its MORPHOLOGY. We can think of something like that in English, though, albeit on a very rudimentary level, and only vestigial, when we think of how the difference between whether a certain lexeme is used as a noun or verb is where the stress hits it: atTRIBute (verb) vs. ATtribute (noun). Don’t worry if that is not how you pronounce them; this is vestigial, like I said.

work in progress

work in progress

bla bla bla

work in progress

we can put the very componants of grammar into a hierarchical ladder:

the most fundamental being the all-permeating phonology; and then comes morphology; and then syntax. Syntax, being the highest, can make use of the lower two. Morphology, likewise, can make use of phonology, but not syntax. Phonology itself is the sort of “ground of being” of Language.

Semantics and the lexicon are left out of this ordering because semantics is implicit to the lexicon, as well as thought itself. It is this semantic thoughtfulness coupled with phonogy that is the basis of language... yet I would not want to put semantics as the base, the lowest rung, with phonology, because sounds of a language can indeed be arbitrary, to refer to Saussue, and do not have to have any meaning in themselves. In English, for instance, we do not ascribe a meaning to the sound represented by the letter B. We could, but we don’t. That is my spin on Sausure - more to say about that in the conclusion. For his simili of Language being like a chessgame, where what move to make next is independent on moves previously made - one just studies the board as is - is really only an admission of what a human mind is capable of knowing. A person can know at once only a snippet of the world around us. One cannot fit in the whole world into one’s head any more than by undoing your bellybutton you youself would fall into it and disappear POOF into thin air!!! we have concepts - snapshots, that is; and that’s all we have to deal with with our finite minds. Shadows in the cave, to allude to Plato. And yet... and yet... The mind is no longer finite when it is no longer tying itseld down with the finite, i.e, concpets... there is a way to let go, to flow with “it”... the unknowable It... unknowable in the sense of not graspable; but known in the sense that one IS it...

more to come! this is a real work in progress with philosophy coming together with linguistics!