Learning a Language is Learning the Structure Signals

003. It is commonly supposed that the major task in learning a new language is learning the lexical meanings of words. This is not in fact the case. The more important as well as the more difficult task is learning the grammatical structures and structure signals (Gleason: 98).

The basic significance of grammatical structures and structure signals in relation to major vocabulary is illustrated by the following text, which has been stripped of all major vocabulary.

In his _____ _____, [Socrates] was _____-ed in the
_____________ ____________ of his _______ and was ___________-ed with
___________ the _________._________ he __________-ed the ________
___________ _________ of ___________ and ________-ing. He ____-ed
the _______ with whom he _______ in _____ about the _____ _____
of ___. He was a ___ of _____ ____ and _____ _____ of
________. He was ______ _______ to ________ and
_____. In ________ he was ____-ed by a ______ _____ on the
________ of ________-ing _______ ______ and of _______-ing the
____. He was ______-ed ____ to _____. He _____-ed to _____.
_____ _____ after his _____ he _____ the _____.

The structure of sentences and word groups is clearly discernible, although the major lexical items have been omitted, with the exception of the subject of the paragraph, Socrates. For example, the sentence beginning in line 4 He ________-ed the _________ is evidently composed of the “subject” he, a verb in the past tense (indicated by the morpheme -ed), and an “object” (the ______); the object is then expanded by a relative clause, with whom …. Similarly, the native speaker of English knows that the sentence beginning in line 6 is that type of sentence in which the “subject” is identified. Taking He was a _______ as a sentence frame, any number of predicate nouns might be found to fit:

He was a philosopher

He was a Greek

He was a father

The predicate, however, is expanded by a prepositional phrase initiated by of. We therefore have this frame: He was a _______ of ______ _______. One might also fill in this frame at random:

He was a philosopher of great wisdom

He was a Greek of great renown

The choice of vocabulary is of course limited by the frame, i.e. by the grammatical structure. Given the frame, the speaker of English knows that certain possibilities are “ungrammatical”:

x He was a good

x He was a very

x He was a philosopher of extremely wisdom

This is to say that the native user of English instinctively knows which vocabulary items are possible choices for each slot.

The grammatical frame of reference is the system within which major vocabulary can be distributed in accordance with rules governing the filling of particular slots or the structure of phrases and the structure of the sentence. Violation of these rules produces phrases and sentences that are grammatically unacceptable or nonsensical.

The prior requirement for the use of a language is knowledge of the grammatical system, i.e. of the grammatical structures and structure signals. This means, with reference to the emasculated text given above, that one must be able to “read” the grammatical frame of reference before he can make any intelligent use of major vocabulary. This appears to be proved by the way in which we actually learn to read and understand languages. We know that a reader can make some sense of a text even when he doesn’t know all the major vocabulary, but that he can make no sense of a text for which the structure signals are missing or jumbled. He was a __________ of great wisdom makes some sense, even when the reader does not know the meaning of the word philosopher, represented here by a blank. But the reader can do little with a sentence like x Was philosopher of Socrates great a wisdom because the structure signals are jumbled. He might put a question mark at the end because he knows questions are often introduced by the verb (in contrast to assertions).

In reading a language we already know, we are often compelled by limited vocabulary to leave certain blanks in the text. We either fill them in by the general sense of the text, or by looking them up in a dictionary. Suppose one wishes to read the paragraph about Socrates but does not know all the words. Twenty-two vocabulary items are omitted in this version:

In his (1) life Socrates was (2)-ed in the (3) philosophy of his time and was (4)-ed with Archelaus the (5). Later he developed the (6) Socratic method of inquiry and (7)-ing He (8)-ed the people with whom he came in (9) about the right (10) of life. He was a man of (11) physique and great powers of (12). He was (13) indifferent to comfort and (14). In 399 B.C. he was tried by a popular (15) on the charge of introducing strange (16) and of (17)-ing the youth. He was (18)-ed to death. He (19)-ed to escape. (20) days after his (21) he drank the (22).

Although many of the words are missing, it is still possible to gain certain information from the text and guess at still more. If, however, twenty-two words like in, was, and, the, a, he were omitted instead of major vocabulary, how much more difficult it would be to make sense of the whole! Consequently, while we can get along in a language with a limited vocabulary, we cannot get along without knowledge of grammatical structure.

The complete text from which the two versions above were drawn reads as follows (it should be noted that not all the lexical choices are necessarily the best ones; an authority on Socrates might well prefer other major vocabulary than the words supplied):

In his early life Socrates was interested in the scientific philosophy of his time and was associated with Archelaus the physicist. Later he developed the famous Socratic method of inquiry and teaching. He questioned the people with whom he came in contact about the right conduct of life. He was a man of strong physique and great powers of endurance. He was completely indifferent to comfort and luxury. In 399 B.C. he was tried by a popular jury on the charge of introducing strange gods and of corrupting the youth. He was condemned to death. He refused to escape. Thirty days after his condemnation he drank the hemlock.

004. The role played by structure signals in the grammatical system of a language can be exemplifed in yet another way, namely by means of nonsense language. Nonsense language is created by employing lexically empty major vocabulary within the framework of the grammatical devices or structure signals of the language. In the often cited English nonsense sentence (Fries: 71)

(i) Woggles ugged diggles

the signals -s, -ed, -s, in that sequence, indicate that the first and third words are “thing words and plural, that the middle term is an “action” word referring to past time. The syntactical relations among the three are also indicated by the sequence: Woggles must be the “subject” of the verb, ugged, and diggles the “object.” On the basis of the structure signals of English and without knowledge of the lexical meaning of a single word, it is possible to construct other “grammatically correct” sentences with these same terms:

(ii) Woggles ugg diggles
(iii) A woggle uggs a diggle
(iv) Woggles will ugg diggles
(v) Diggles have been ugged by woggles
(vi) The woggles which ugged diggles were ugging woggles

In our nonsense language use has been made of only two “thing” words and one “action” word, plus the structural items of standard English. We could enlarge the scope of our language by creating one new word:

(vii) Waggles are fabothful

Our knowledge of English tells us that the morpheme -ful identifies the new word as a descriptive word; this is confirmed by its position after are in the sentence. On this basis we can now create an additional sentence:

(viii) Waggles ugg fabothfully

and then expand it:

(ix) The fabothful woggles ugged diggles fabothfully.

(ix) shows that fabothful may “modify” a “thing” word, and fabothfully an “action” word.

If we add the English structure word and to our repertoire, it is possible to introduce other structures, e.g.

(x) Woggles and diggles have been ugged

in which two “subjects” are linked to a single verb. And may also link two sentences:

(xi) Woggles ugg diggles and diggles ugg woggles

Nonsense sentences (i)–(xi) contain at least four sentence structures basic to English:

(a) Woggles ugg fabothfully (viii)

which may be transformed into a sentence with English lexical items:

  Altos sing beautifully  
(b) Woggles are fabothful (vii)
  Sunsets are beautiful  
(c) Woggles ugged diggles (i)
  Teachers harassed students  
(d) Diggles have been ugged by waggles (v)
  Students have been harassed by teachers  

English nonsense sentences demonstrate that grammatical structure in some important respects is independent of the lexical meaning of major vocabulary items. One does not learn the meaningful arrangement of words by learning major vocabulary. Teacher the students harassed the contains standard English vocabulary but is only a jumble of words. It is this important fact that justifies the assertion that in learning a language the prior and more important task is learning the structures and structure signals.

005.1 In the nonsense sentence, Woggles ugged diggles, -s, -ed, and -s are morphological variables providing clues to the structure of the sentence. They are called morphological variables because they have to do with the change in the form of words in contrast to some other possible form (woggles is shown by -s to be plural, in contrast to the singular form woggle, without -s).

005.2 In the modern printed Greek text the discrete items on the page are (for the most part) words, e.g. πάλιν, αὐτοῖς, ἐλάλησεν. Some of these never vary in form, e.g. πάλιν οὖν ἐν. Others vary in form in accordance with their function within the structure of the word cluster or sentence, e.g. αὐτοῖς, ἐλάλησεν. Morphological variables occurring in words therefore indicate word relationships and participate, along with other features of the language, in the larger structure or system known as grammar.

005.3 The signals ἐ----σεν (ἐλάλησεν Jn 8:12) indicate that the word in question is verb, that it refers to past time, that the “subject” is singular and acting rather than being acted upon. On this basis one can readily identify the following: ἐ-πία-σεν (Jn 8:20), ἐ-ποίη-σεν (Jn 8:40), ἐ-τύφλω-σεν (1 Jn 2:11). The signals τοῦ----ου (τοῦ κόσμου Jn 8:12) suggest that the word in question is a “thing” word, that it is singular in number, and that it stands in a “genitive” (i.e. “of,” see §112.1) relation to some other word or word group in the sentence. Cf. τοῦ ἀνθρώπου (Jn 8:28), τοῦ θεοῦ (Jn 8:40), τοῦ ἱεροῦ (Jn 8:59).

005.4 The Greek sentence, like the English sentence, exhibits certain structure signals in the form of morphological variables that the competent reader of Greek is able to “read” at a glance. In the sentence

(1) ἡ σκοτία ἐτύφλωσεν τοὺς ὀφθαλμοὺς αὐτοῦ 1 Jn 2:11

the signals ἡ ----α and τοὺς ----οὺς indicate that the items in question are “thing” words, in the first instance singular, in the second plural. The signals ἐ-----σεν indicate the word in question is a verb, with reference to past time. ἡ -----α indicates, furthermore, that the first word group is “subject,” τοὺς -----οὺς that the third word group is “object,” of the verb. The signal —ου, as part of the last word, indicates that this word stands in a “genitive” relation to some other item in the sentence.

005.5 The reader will grasp (1) as a whole consisting of three parts, probably of this order:

ἡ σκοτία / ἐτύφλωσεν / τοὺς ὀφθαλμοὺς αὐτοῦ.

He will be led to grasp the first and third word clusters as units because of the sequence of morphological variables, ἡ -α go together as feminine article (the) and feminine noun, nominative singular. τοὺς -οὺς go together as masculine article and masculine noun, accusative plural. He also knows that nominal word clusters commonly embrace a "modifying genitive," so that αὐτοῦ is taken with the preceding cluster.

005.6 Position (word order) also plays a role in seizing word groups, e.g. the words in the first and third clusters occur together and in some acceptable order: the position of article (ἡ, τούς) and noun (σκοτία, ὀφϑαλμούς) in relation to each other is fixed — the article always precedes; the genitive αὐτοῦ (pronoun), however, may also come between or before article and noun (and occasionally may be separated from its word group). Word order plays a significant but quite different role in Greek sentence structure than in English.

006.1There are other structure signals in Greek, as in English, in addition to the morphological variable and word order. Another group of signals consists of words, some of which are fixed in form (indeclinables), and some of which vary in form (inflected). This category is called function words (cf. §002). In the case of inflected function words (cf. §002). In the case of inflected function words, the morphological variable(s) and the word as a whole, as signals of different orders, converge.

006.2 Questions in Greek are often signaled by interrogative words. In the sentence

(2) ποῦ ἐστιν ὁ πατήρ σου; Jn 8:19
Where is your father?

ποῦ is an interrogative adverb corresponding to English where?. It is fixed in form. On the other hand, τίς is an interrogative pronominal adjective that is inflected. In the sentence

(3) τίς ἐστιν ἡ μήτηρ μου; Mk 3:33
Who is my mother?

τίς not only signals a question, but its role in the structure of the sentence is indicated by its case form (nominative); it is "subject" of the verb ἐστιν.

006.3 Some function words typically signal subordinate clauses (included dependent sentences).

(4) ταῦτα λέγω ἵνα ὑμεῖς σωϑῆτε Jn 5:34
I say these things in order that you may be saved

In (4) ἵνα indicates that a final (purpose) clause is being joined to the main sentence. Similarly, εἰ regularly introduces a subordinate clause which serves as the protasis (if-clause) of a conditional sentence:

(5) εἰ ἐμὲ ᾔδειτε, καὶ τὸν πατέρα μου ἂν ᾔδειτε Jn 8:19
If you had known me, you would have known my father also
Such function words as ἵνα and εἰ are known as subordinators.

006.4 Another subclass of function words is known as conjunctions. They are employed to connect or relate sentence elements of the same or similar order. They may be single words, as in

(6) ἐπίστευσεν αὐτὸς καὶ ἡ οἰκία ὅλη Jn 4:53
He and his whole house believed

where καί links two subjects with a single verb; its grammatical function therefore corresponds to and in the translation, καί is thus an essential clue in the structure of (6). Two such conjunctions may be correlated:

(7) οὔτε ἐμὲ οἴδατε οὔτε τὸν πατέρα μου Jn 8:19
You know neither me nor my father

οὔτε ... οὔτε links two objects to a single verb and thus corresponds to neither ... nor in the translation.

006.5 ἴνα, εί, καί, οὔτε are fixed in form, but not all function words in these subclasses are indeclinable.

007.1 Derivational forms are of assistance in assigning words to word classes. The word sad in English is an adjective, as illustrated by the sentence, that face is sad, to which the phrase, that sad face, is related. By adding the suffix -ly, sad is turned into an adverb: sadly. The suffix -ness transforms the same word into a noun: sadness. The class membership of derivational forms is grammatically significant. The native user of English knows that the sentence x that face is sadly, or the phrase xthat sadly face, is grammatically unacceptable. The matter can be rectified by adding or inserting an adjective:


That face is sadly wrinkled

That sadly wrinkled face.

Or again, sadness fits certain sentence frames that sad and sadly do not. For example, we may say


He had a profound sense of sadness

but not:


xHe had a profound sense of sad

xHe had a profound sense of sadly
xThat sadness face.

The derivational pattern of sad, sadly, sadness is sufficiently common in English as to suggest other analogous sets:


glad, gladly, gladness

mad, madly, madness
bad, badly, badness

On the basis of the derivational forms the class membership of these terms is immediately evident to the speaker of English, in practice if not formally.

007.2 Derivational forms are also of use in identifying the class membership of words in Greek. A pattern not unlike that illustrated in §007.1 is exemplified by

  δίκαιος righteous (adjective)
  δικαιοσύνη righteousness (noun)
  δικαίως justly (adverb)

Other nouns may be developed from the same stem:

  δικαίωμα righteous act
  δικαίωσις justification

-συνη, -μα, -σις are noun suffixes, just as -ως is an adverb suffix.

007.3 On the basis of these derivational suffixes, the following words can be identified as nouns:

  ἐλεημοσύνη charity
  ὄνομα name
  πνεύμα spirit
  γνώσις knowledge

And the following can be identified as adverbs:

  ἀληϑῶς truly
  ταχέως quickly

The examples of nouns and adverbs listed above do not necessarily follow the pattern of δίκαιος in other respects.

008. The structure-signaling devices of Greek form a coherent and finite system. Because the system is coherent, the structure signals of any phrase, sentence or series of sentences are to be read as a constellation of signals pointing to one grammatical structure or set of structures, in distinction from other possible structures or set of structures. If the structure is "grammatical," the signals will converge in pointing to that structure. That is only to say that "grammaticality" and coherence of the structure-signaling system are one and the same thing. It goes without saying that no grammatical system is perfect, i.e. free of ambiguity.

Because the grammatical system is finite it can be learned without having to learn all possible Greek sentences. The potential output of Greek sentences is infinite, but the grammatical system is sufficiently restricted so that many men, women, and children — not all of whom were bright or linguistically gifted — were able to learn to read and speak Greek.

The Aims of Descriptive Grammar

009. A descriptive grammar must achieve two things: 1) it must develop a classification of words and other elements; 2) it must state the grammatical relations that obtain among these elements (Gleason: 138).

The two aims of descriptive grammar may be modestly elaborated, with particular reference to the shape of the present grammar. The balance of the introduction will thus be an immediate introduction to and sketch of the content of the grammar.