Apollonius and Priscian, the great grammarians among the ancients
Apollonius
In Plato’s time a philologist was a cultured man; in Hellenistic
times (the period of Hell?nismos) the philologist was a student and expositor
of literature and correct standards in language use with the aim to imitate –
as well as to fully understand and appreciate – the language and style of the
classical Athenian writers.
We saw the intention of the Alexandrian grammarians
to develop literary style and an appreciation of the classical literature in
the Techn? Grammatik? attributed to Dionysius Thrax. The TG contains nothing on
the combination of the parts of speech into sentences but it did, along with
other works now lost, form the basis for later writings on grammar and syntax.
None achieved the authority of the grammar of Apollonius Dyscolus. Apollonius
(c. 80–160 CE) lived all his life in Alexandria. His works reveal that he was
dismissive of opposing views, often describing them as silly or absurd or
stubborn, e.g. The people who specify that the Dorians don’t form contracted
future subjunctives, then, and investigate the question why they don’t are
plain silly. […] And if those who listen to the complete argument refuse to
share this opinion [of mine] they are behaving extremely stubbornly.
(Apollonius Dyscolus 1981 III: 141)1 Also his writing is dense and sometimes
difficult to follow. When not quoting one or another of the literary giants of
ancient Greece, the language examples he invents often suggest the schoolroom;
so he was probably an impecunious teacher.2 Certainly he had a reputation for
being ill-tempered, hence the sobriquet Dyscolus “grouch”. Apollonius was not
the first to write on syntax, though little or nothing survives of his
predecessors’ works; but in any case he appears to have written the most
comprehensive account.
The extant works of Apollonius are: On Syntax, On the
Pronoun, On Adverbs, On Conjunctions, Fragments (Parts of Speech). 3 His work
takes up that of scholars like Dionysius Thrax on parts of speech, even though
Dionysius himself is not named at all in On Syntax (henceforth Synt.). By
contrast, Tryphon (roughly contemporary with Christ) rates 52 mentions (and
Tryphon’s student Habro, nine), Aristarchus (c. 217–145 BCE) is mentioned 24
times, and Zenodotus (third century BCE) 14 times (Householder 1981: 5).
Apollonius’ work shows clear evidence of influence from the Stoic grammarians:
for instance, that a sentence is complete only if it has a definite subject
(Synt. I: 14); that proper names express individual quality (idia poiot?s)
(Synt. I: 78); and 1. Translation from the Greek is based on Householder’s
translation in Apollonius Dyscolus 1981, with occasional modifications. A
French translation by Jean Lallot alongside the Greek is Apollonius Dyscole 1997.
2. Teaching was badly paid throughout the ancient world, too (Marrou 1956:
204). 3. Peri Suntaxe?s Uhlig (ed.) 1883 II.ii; Peri Ant?numias II.i.3–116;
Peri Epirr?mat?n II.i.117– 210 [sic]; Peri Sundesm?n II.i.211–58; Fragments
II.iii. 102 The Western Classical Tradition in Linguistics The construction
[parathesis] of a nominative with a transitive verb always remains half
complete [because it requires an object NP]: Truph?n blaptei [“Tryphon harms”]
or Truph?n philei [“Tryphon likes”]. That is why the Stoics gave the name
‘partial-predications’ [elatton ? kat?gor?mata] to such verbs in contrast to
the [intransitive] verbs that are complete in themselves and do not always
demand an oblique case complement. (III: 155) Apollonius’ writings (especially
Synt.), along with lost work of his son Herodian (Herodianus), formed the basis
for Priscian’s study of Latin syntax. Priscian several times described
Apollonius as the greatest authority on grammar (Priscian 527, VI: 1; VIII: 87;
XIV: 1; XVII: 1),4 adopted his methods and, where applicable, his analysis. So
Apollonius Dyscolus was a powerful influence on the Western Classical Tradition
in linguistics. Apollonius believed that linguistic phenomena are ordered
according to knowable rational rules of grammar; knowing these rules is a
better guide to correct language understanding and production than merely
observing the usage of either poets and dramatists or ordinary speakers. There
is a precursor here of the Chomsky 1965 distinction between competence and
performance: in the view of Apollonius, authorities may disagree on what counts
as proper usage and everyone has gaps in their knowledge. It is knowledge of
the rational analysis that provides a firm basis for decision and allows one to
spot errors in usage. Just as the utility of the literary tradition is very
great for correcting both the texts of poems and the usage of everyday speech,
and determining the application of words among classical authors as well, in
the same way also our present investigation of grammaticality will provide a
rational correction for all sorts of errors. (I: 60) Apollonius correlated
linguistic explanation with an understanding of different levels of adequacy in
grammar. Towards the end of Book III he writes: The foregoing account [of noun–verb
relations] is adequate for anyone who wishes merely to understand the facts of
usage and tradition [paradosis]. But for anyone who wants to investigate in
detail the underlying theory of grammar [ta t?s suntaxe?s tou logou], it will
be necessary to inquire which verbs take the genitive and why, which the
dative, and again for what reason, and finally also the accusative. Obviously
it is going to be a big and difficult task. (III: 158) Notwithstanding his
theoretical bias, Apollonius requires that all grammatical claims be data-based
and apply to all the data: I rely not merely on poetical citations, because
poetic constructions can be either elliptical or pleonastic, but on common
everyday usage, the practice of the best prose-writers, and, most of all, on
the force of theory which must be applied even about constructions which are
not in the slightest doubt. (II: 49) Meanings rely on contextual information:
we must always determine the sense not by studying the accent, but from the
sentence context, just as, in other types of indefinite ambiguity, distinctions
are made from the sentence context, not from the presence of enclisis
[inflection] or accent. (II: 102) 4. E.g. ‘Maximis auctoribus Herodiano et
Apollonio’, Inst. VI: 1; ‘Apollonius, summus artis auctor grammaticae’, Inst.
VIII: 87. Apollonius and Priscian 103 Synt. does not examine all the rules of
grammar but only those which raise interesting questions to serve as exemplars
of the method to be used in linguistic analysis. Apollonius adopted the Stoic
view that there is a naturally correct way of expressing every thought clearly,
unambiguously, and correctly; idiosyncrasies and anomalies are perversions of
natural expressions. Once irregularity is understood and explained it is no
longer a threat to the regularity within the system. This idea seems to have
been adopted by Priscian: So then, from the ordering of words we determine the
sense of the construction, whether it be correct or not. For, if it be
anomalous, it will create an error. (Priscian 527, XVII: 6)
The sequence for
presenting a syntactic analysis follows the supposed order in which the various
parts of speech were brought into being. It was believed that the sequence of
letters in the alphabet, ????????, etc. is natural. The natural order of the
parts of speech, the sequence in which they were assumed to have been
phylogenetically invented and ontogenetically acquired,5 is noun, verb,
participle, article, pronoun, preposition, adverb, conjunction. [T]he noun is
prior to the verb. (Synt. II: 4) The noun necessarily precedes the verb, since
influencing and being influenced are properties of physical things, and things
are what nouns apply to, and to things belong the special features of verbs,
namely doing and experiencing. (I: 16) The natural sequence of cases is
nominative, genitive, dative, accusative, vocative; for genders masculine,
feminine, neuter (Synt. I: 13); for tenses it is present, imperfect, past,
future (‘the present and past are known but the future is uncertain’ I: 114);6
for mood7 we begin necessarily with the indicative mood, not because it is
indeed primary, but because it is the most transparent, occurs frequently, and
can provide instructive cases of homophony, phonological changes [pathos] and
derivation. (III: 62) In general, each definition only has to mention
previously defined classes.
In Book III: 13, Apollonius divides the parts of
speech into the four superclasses of Table 6.1. The proper inclusion of the
word classes of C within A suggests corruption of the text. Certainly Priscian
527, XVII: 153 combines the overlapping C with A and replaces it with a
temporal category (Table 6.2); this may have been innovative or a faithful
reproduction of Apollonius’ original scheme.
Tables 6.1 and 6.2 should be
compared with Varro’s use of ±case, ±tense to categorize the parts of speech
(presented in Table 4.3). Apollonius makes no significant use of the
superclasses in Table 6.1. The four books into which Synt. is divided are
slightly more closely aligned to Priscian’s superclasses in Table 6.2: Book I
is on 5. One can attach some credibility (not much) to the naturalness of the
supposed sequence of the parts of speech, see above p. 89; but absolutely none
for the naturalness of the sequence of letters in the alphabet. In Synt. I: 18
Apollonius says that the name alpha comes from alphein “to know one’s ABCs”
(Coogan 1990): ‘A took the name of all the letters to itself because of its
initial position, and the coincidence of the initial sound of the word.’ The
very opposite is more likely to be true. 6. Apollonius has no distinct name for
aspect, referring to both the perfect and the aorist as past. 7. See p. 90
above. Apollonius on mood is discussed in more detail at pp. 109f. 104 The
Western Classical Tradition in Linguistics Table 6.1. Apollonius’ four
categories among the parts of speech. A. +case, +number noun, participle,
article, pronoun B. +person, +number verb, pronoun C. + gender noun, article D.
–case, –number, –person, –gender preposition, adverb, conjunction Table 6.2.
Priscian’s four categories among the parts of speech. A. +case, +number, +
gender noun, pronoun, participle B. +person, +number verb, pronoun C. +
tense/time verb, participle D. –case, –number, –person, –gender, –tense
preposition, adverb, conjunction, interjection the definite article and the
relative pronoun;8 Book II on pronouns; Book III on solecism (syntactic error),
verbs and their arguments and participles; while Book IV deals with
prepositions, adverbs and conjunctions. On Syntax (Synt.) begins In our
previous publications we discussed the theory of words [ph?n?], as the nature
of the subject required; the present work will cover the topic of the
combination [suntaxis] of these words, according to appropriate collocation
rules, into independent sentences, a topic which we have chosen as deserving
the greatest precision because of its essential importance for the
interpretation of poetry. (I: 1) Apollonius pointed out (I: 8) that one must
recognize correct spelling in order to recognize correct structure; that the
letter (stoicheion) is unanalysable; and that syllables are structured into
words (lexis). [W]ords […] being the primes of a regularly constructed complete
sentence [logos], must accept restriction by the structural rules of syntax.
For the meaning [no?ton] which subsists in each word is, in a sense, the
minimal unit [stoicheion] of the sentence, and just as the minimal units of
sound compose syllables when properly linked, so, in turn, the structural
combining of meanings will produce sentences by combining words. Just as the
word is made of syllables, so the complete sentence is made by grammatical
collocation of meanings. (I: 2) So he has grounds for claiming (I: 9) that If
Dionysius is walking he is moving is true but the converse is false. For
Apollonius the word is the minimal unit of grammatical analysis; he has no
conception of ‘morpheme’ (thus when I use the term ‘morphosyntactic’ in
describing his work I am imposing a modern concept that applies to what he is
discussing).
Apollonius used some of the same criteria in defining his parts of
speech as Dionysius Thrax but he gives far more detail and combines semantic
with formal explanation. For the phonological forms of words do not weigh as
much in classification as what is meant by them. (II: 33) 8. The article class
includes the relative pronoun in Greek due to morphological similarities in
form and behaviour: relative pronouns normally follow the noun and agree with
it. Colson 1919: 24f suggests that ‘[the article and the pronoun] are the
counterparts of one another, dealing with the meaning of the word as a single
entity.’ Apollonius and Priscian 105
Showing his Stoic credentials, Apollonius
was responsible for offering a common class meaning for each of the eight parts
of speech, and so for the purely notional descriptions that came to
characterize parts of speech in the Western Classical Tradition.
Though he gave
priority to meaning he did not ignore morphosyntactic form and function. He
distinguished between schema (form) and ennoia (meaning) and assigned
grammatical structure to the category of ennoia. Pinborg 1975: 106 claims that
it was ‘through the work of Apollonius and his son Herodian, that uniformity
[was first] achieved.’ The degree of achievement depends on whether the formalized
word class system in our text of the Techn? Grammatik? was available to him
from whatever source; but this is not to doubt that Apollonius himself was
perfectly capable of devising such a system. He correctly recognized (Synt. I:
37) that the mention of a word (e.g. the conjunction men) presupposes knowledge
of its morphosyntactic classification. Apollonius defined the pronoun not only
as noun substitute, as the Techn? Grammatik? does, but in addition as referring
to substance without qualities (i.e. pronouns are semantically minimalist) – a
statement that Priscian repeated and which was later used in medieval grammars.
Focusing on their anaphoric function, he said that pronouns ‘are used where
nouns cannot be used or where they have already been said once and cannot be
used again’ for stylistic reasons (Synt. II: 11). Because word boundaries were
often not marked in Ancient Greek texts, this matter is a focus for Apollonius
when he discusses whether prepositions/prefixes are separate words or not. Our
distinct terms preposition for the separate words and prefix for affixations
help to obviate the difficulty. Apollonius’ use of analogical, distributional,
and phonological criteria (such as accent placement) is similar to methods used
today. This preposition [or prefix] will not gravitate toward any other word
than one that can accept a relationship with it. And so *emPlat?n is impossible
but embainei [“comes in”] is all right; likewise *sumPlat?n is out but
sumbainei [“comes along”] is OK, as are diabainei [“crosses”] and peribainei
[“comes around”]. Since such a sentence has only one person altogether – I mean
sentences like Plat?n bainei – when a preposition is brought in from outside it
will gravitate to the position before the verb, since it is a prepositive
element and the verb is postpositive. (IV: 15) Apollonius had no terms for
phrase categories such as noun phrase, verb phrase, prepositional phrase, or
adverbial phrase, etc. He had no term for “clause” (which is at best logos
“sentence”). Nor did he have names for grammatical relations like subject,
direct object, indirect object; instead he uses the case terms nominative for
subject, accusative for object, etc. Very occasionally he uses topic where we
might use “subject”. Other terms and notions he employs are: sunthesis A
compound: usually two words put together to form one lexeme. parathesis A
construction: overlaps with our notion of phrase but is not identical to it.
suntaxis A construction with a syntactic relationship between the component words
as in onomatik?/epirr?matik? suntaxis “nominal/adverbial construction” (Synt.
I: 55). Both parathesis and suntaxis are constructions but they are never
treated as constituents within larger constructions. Both terms occur in
parathesis prothetik? kai arthrik? suntaxis “prepositional placement and
articular construction” i.e. preposition followed by relative pronoun (classed
as an article) as in eis ho “to which” (IV: 57) (Householder 1981: 2). 106 The
Western Classical Tradition in Linguistics Apollonius frequently uses the term
katall?lon or katall?lot?s, which is concord among word forms. Katall?lot?s is
a measure of the regularity and rationality of a construction; the grammatical
correctness will be in agreement with the meaning. Irregularity is the result
of corruption. Apollonius accurately and graphically distinguishes
inappropriate reference from solecism: [W]hen someone says houtos me etupsen
[“this.M beat.3.SG me”] of a woman, it is not a grammatical error. The sentence
obeys all the rules of agreement and government [katall?lon]. But if, referring
to a female, someone were to say *haut? me etupsan [“this.F beat.3.PL me”],
even if the gender reference is correct, this is a solecism because of the
error in [number] agreement. (III: 10) Correct or incorrect reference has
nothing to do with katall?lot?s. On the other hand: The masculine of husterik?
[“suffering in the womb”] or of ektrousa [“having miscarried”] are
morphologically possible but semantically not. (III: 149) which is the other
side of the coin. Apollonius was sensitive to scope relations, recognizing that
ho deipn?sas pais koimasth? is ambiguous between “Let any boy who has dined go
to bed” and “Let the particular boy who has dined go to bed”, the latter being
similar to ho pais deipn?sas koimasth? (I: 111). He also demonstrates that the
article ‘belongs to the thing possessed’ (I: 101f); thus although both ho emos
pat?r (the my father) “my father” and emos ho pat?r “the father is mine” are
possible, *ho emos ho pat?r philosophei is not; it must be ho pat?r ho emos
philosophei (the father the my philosopher) “my father is a philosopher”.
Apollonius based his description on the relations between nouns and verbs, and
he noted how these relations were revealed by cases. Instead of identifying
grammatical relations, Apollonius distinguishes two ‘persons’ (pros?pa): one
diatithen “affecting” or energoun “acting” – perhaps both best translated actor
in the sense of Role and Reference Grammar;9 and the other diatithemenon
“affected” or energoumenon “acted upon”, i.e. undergoer. All datives other than
instrumentals and temporals are benefactive, therefore he had no need to
identify an indirect object. His treatment of verb syntax suggests a rule
something like (1) with an implicit rule putting the nominative NP before V;
e.g. from (2) derive (3). (1) S ? V (NP)n (2) S ? V NPNOM (NP)n (3) NPNOM V
(NP)n He also recognizes NP sequences such as ART?ADJ?N?REL. According to
Householder 1981: 17 (thinking in terms of post-Chomsky 1965 grammar),
‘Apollonius Dyscolus was the inventor of the abstract base.’ Synt. apparently
justifies this claim. It is certainly true that Apollonius proposes underlying
constructions that explicitly represent meaning by using constructions that are
in various ways distinct from the normal surface sentences of Greek; his
underlying constructions need to be transformed into surface structure by
epenthesis, rearrangement (huperbaton), ellipsis and substitution (enallag?).
The use of such transformations in etymologies presumably predates Plato; but
Apollonius seems to have developed the concept. And some of his proposed
underlying structures bear 9. See Van Valin and LaPolla 1997; Allan 2001; and
Chapter 11 below. Apollonius and Priscian 107 a close resemblance to analyses
independently proposed in the late twentieth century. For Apollonius the
underlying structures are the regular and correct grammar from which both
ordinary speech and poetry derive; though poets typically make more
transformations than the ordinary language user. For instance, he insists that
the protasis of a conditional (sunaptikos) logically precedes the apodosis;
thus conditional ei “if” and para-conditional epei “since, when”, etc. are
logically sentence initial even in (4) ph?s estin ei h?mera estin light is if
day is “It’s day if it is light” For it is only in the surface order
[sunthesis] that ei h?mera estin [in (4)] stands second, not in the explicit
semantic structure [diexodik? epangelia]. For the mind must first accept the
thought it is day, and then it can accept it is light. So never can we agree
that the conditional particle ei is not sentence initial [semantically] even
though it may sometimes occur later in the sentence. (II: 77) On underlying
case relations: In effect one who says ton emon doulon epaisa [“I struck my.ACC
slave”] is saying *ton emou [GEN] doulon epaisa which must become ton emautou
[1.SG.GEN.REFLEXIVE “my own”]. (II: 106) In Ancient Greek, verbs of ruling take
the genitive and according to Apollonius (III: 175f) the verbs are denominal,
e.g. turann? hum?n [1.SG.rule you.GEN “I rule over you”] and turannos hum?n
[ruler.GEN you.GEN, “your ruler”]. Thus although in Aristarchou doulos
[Aristarchus.GEN slave.NOM] the possessor head noun is genitive, in kurieu? eg?
tout?n [1.SG.be.lord.of I.NOM them.GEN] the possessor is nominative and the
possessed genitive. The verb requires a nominative subject and possession is
dependent on that. He also explained the fact that philein “to love” takes the
accusative while eran “to love” takes the genitive on the ground that the
latter involves more passionate love (compare English philology versus
eroticism). Turning to the dative: in leg? soi [I.tell you.DAT] the dative is
correctly used because ‘I am benefiting you’ but in leg? se klept?n [I.tell
you.ACC thief, “I call you a thief”] signifies [s?mainei] I assert by means of
the speech which I am uttering that you have done the act of stealing […
because it] directs the act against the object and so takes the accusative
construction. (III: 177) So-called ‘psych verbs’ take a dative subject, e.g.
melei/metamelei Sokratei (cares/repents Socrates.DAT) “Socrates cares/repents”.
the Stoics called these verbs pseudo-predicates [parasumbama] whereas the other
verbs, according to their subject-verb relations, were called by them either
sumbama or kat?gor?ma [“predicate”]. (III: 187) On reflexives: Anyone who says
emauton etupsa [“I hit myself”] affirms a relationship coming from himself to
himself, as if he were saying eg? eme auton etupsa [“I myself hit me myself”].
(II: 139) Boulomai ploutein “I want to be rich” implies a reflexive: boulomai
ploutein emauton “I want myself to be rich” (III: 162). 108 The Western
Classical Tradition in Linguistics Sull?psis is the amalgamation of several
notions into a portmanteau lexeme, e.g. fiercer is analysed as “more fierce”
and a stable as “a barn containing horses” (III: 61). Similarly, Tryphon spoke
to me and you and Dionysius can be expressed as Tryphon spoke to us where ‘us’
is an example of syllepsis (III: 38). We also know that every plural form of
first and second person involves a conjoining of different persons: h?m?s
[1.PL] is a reduced expression for either me and thee and him or me and you or
me and them. Similarly in the case of the second person: hum?s [2.PL] is thee
and them. (II: 159) Time adverbs like echthes “yesterday” and aurion “tomorrow”
have no gender but in construction with prepositions they take the feminine
article: Apollonius assumes that there is an implicit feminine noun h?mer?
“day” in expressions like en t?i echthes (during the.F yesterday) “during
yesterday” and en t?i aurion “during tomorrow” (IV: 70). Explicating the
meaning of the conditional conjunction, Apollonius is sorely in need of a term
for aspect or aktionsart, which he does not have: e?n math? [“if I learn”]
which means ei anasumai to mathein [“if I finish the act of learning”], or ean
dram? “[if I run” means] ei anasumai to dramein [“If I finish the act of
running”]; whereas ean trech? means ean en pararesei gen?mai tou trechein [“if
I get into the continued process of running” …] the conjunctions themselves
signify the potential future, whether progressive [paratasis] or perfective
[anusis]. (III: 140) This aside, Apollonius makes intriguing analyses of the
verbs and their subcategories of mood, voice, type (primitive or derived), form
(simple, compound or super-compound), number (singular, dual, plural), person,
and tense. We have seen brief glimpses of the considerable attention he devoted
to concord between nominative noun and verb and to the verbal government of
cases. According to Hahn 1951: 31, it was Apollonius who first noted that ‘a
given verb-form is an enklisis; it shows or possesses a diathesis.’ This sums
up concisely what is outlined, but not explained in detail, in the Techn?
Grammatik?. Apollonius argues that person, number, and mood are not properties
inherent in the verb, but expressed by addition (enkliseis) to the verb (I:
51). The term diathesis is used of the three kinds of voice: active, passive
and middle (or neutral). the category of voice […] is present in every mood,
not even excluding the infinitives, because of the logical necessity for all
tenses to be marked as either active or passive or middle. (III: 147) Active
verbs denote an action ‘passing over to something or someone else’; in other
words, the clauses in which they occur are transitive; passive and middle10
verbs occur in intransitive clauses. Apollonius noted the case structure of
nouns for functional relations with these species of verbs. But perhaps the
most interesting insight of Apollonius was recognizing that each mood or
clause-type has a psuchich? diathesis “mental disposition”, which Householder
1981 10. The middle voice focuses on an action of benefit to the subject, e.g.
middle elousam?n “I bathed [myself]” vs. the active elousa “I bathed [someone
else]” (cf. Synt. I: 30). Apollonius and Priscian 109 correctly identifies as
illocutionary force.11 [E]very mood may be paraphrased using an infinitive as
the generic name of the verb. E.g. if we have a declarative sentence peripatei
Truph?n [“Tryphon is walking”] we can turn it into a report of the utterance by
adding the verb implicit in the indicative mood, namely h?risato [“He
declared”], giving h?risato peripatein Truph?na [literally, “he.declared to.walk
Tryphon”]. And similarly for the optative form peripatoi? Truph?n [“May Tryphon
walk”]. Here too one may supply [the verb] inherent in wishing, and say ?uxato
peripatein Truph?na [“He wished for Tryphon to walk”; see also Synt. III: 95].
And similarly for the imperative mood peripateit? Truph?n [“Let Tryphon walk”
one would say prosetaxe peripatein Truph?na [“He ordered Tryphon to walk”]. (I:
51). The Homeric practice of using the infinitive form by hypallage for the
imperative construction, I think, is also explained by its generality, the fact
that all special [moods] can be transformed into infinitives. (III: 63) But
infinitives, since they [unlike finite verbs] have not yet acquired subject
persons, naturally also have no expression of the mental attitudes of those
persons. (III: 59) [Because the infinitive] has no mental attitude
[illocutionary force] it cannot be blocked from occurring with all moods
[enclisis] with the addition of words signifying the characteristic mood; and
conversely every mood can be converted into the infinitive. For graph?
[“Write!”] can be equated to graphein soi prostass? [“I bid you write!”] where
we necessarily add the bid which is implicit in the imperative along with the
pronoun [you]. For the infinitive has no share in either of these [mood or
person]. Peripatoi?s [“May you walk!” is equivalent] to euchomai se peripatein
[“I pray you walk!”], and grapheis [“You are writing”] to horizomai se graphein
[“I declare that you are writing”]. The conversion [metal?psis] is obvious also
[in reports]: graphoi Dionusios [“Would that Dionysius write!”] goes to ?uxato
graphein Dionusion [“He prayed for Dionysius to write”], and graph?to Dionusios
[“Let Dionysius write!”] to prosetaxen graphein Dionusion [“He gave the order
for Dionysius to write”]. (III: 25; see also III: 61) In passages like these
Apollonius foreshadows an analysis that reappeared some 1800 years later in
Ross 1970: 261 as ‘Every deep structure contains one and only one performative
as its highest clause’ causing it to be known as ‘the performative analysis’.12
The postulated deep structure was I ?PERFORMATIVE?you?S, which is not quite the
same as what Apollonius wrote of. For instance, in the quote from I: 51
Apollonius uses a third person subject rather than first person; and he
certainly had a less sophisticated notion of illocutionary force than obtained
in the late twentieth century. Nevertheless his depth of insight is salutary.
It gives him the basis for writing of negation [The indicative mood] contains
the force of affirmation [kataphasis]. And this is why the socalled negative
adverb of denial [ou(k)], which has the force of fighting the yes assertion,
regularly accompanies the indicative mood [horistik? enklisis] in order to
reject the inherent 11. Here ‘illocutionary force’ is what Allan 2006a refers
to as the ‘Primary Illocution’ of the clause-type. There is some overlap
between mood and clause-type and they were mistakenly identified in Allan 2001,
but distinguished in Allan 2006a. 12. See p. 347 on performatives. There is
more detail on the performative analysis and its flaws in Allan 1986: 8.10.3;
Allan 1994, and works cited therein. Allan 2006d argues that Apollonius can be
interpreted to identify the primary illocution of clause-types – though Apollonius
would have spoken of mood. 110 The Western Classical Tradition in Linguistics
affirmation: ou graphei [“Not he is writing” (I deny that he is writing)], ou
peripatei [“Not he is walking” (I deny that he is walking)].’ The imperative
and optative take the prohibitive particle m? instead of ou(k). (III: 90) This
combines the Stoic account of sentence negation, which is the direct ancestor
of modern semantic accounts of negation, with his account of the illocutionary
force implicit in each clause-type. The Syntax of Apollonius Dyscolus is a
magnificent tour de force that argues for the semantic basis of grammar, while
still paying close attention to formal aspects of the objectlanguage.13 We can
look at it as principally an exposition of the language structures found in
classical Greek literature and in first century Greek; the main intention was
apparently to bring students to recognize the regularities that Apollonius
perceived in the morphosyntax of the language and for which he offered rational
explanations. The regularities often underlie a less regular surface structure
which is derived through transformations deleting, inserting, and transposing
the semantically based underlying structures. The extant text of On Syntax is
fairly comprehensive as it stands, and Apollonius’ complete oeuvre certainly
seems to have been comprehensive. It deals with all aspects of literary craft –
poetic effects, tropes, figures of speech, manipulation of syntax – that would
enable his students to understand and emulate the classical literature.
Implicit in Synt. is the view that the classical literature was superior in its
language to contemporary spoken Greek; although Apollonius does not belabour
such a view. Apollonius’ son Herodian is best known for his work on prosody. Later
prosodists described not only pitch levels, vowel length, syllable quantity,
and aspiration but also vowel elision, pitch changes resulting from word
compounding and derivation, and word boundary juncture, e.g. Let’s get away
Bill vs Let’s get a weigh-bill; a whole nother thing; and, for some Australian
speakers, a night vs an eight. In Greece, grammar developed from interest in
the forms of language in which philosophical speculation and logical analysis
were conducted. The use of language in epistemological explanation also
encouraged the investigation of the way in which language operates and, too,
the way in which it originated. The pursuit of Hell?nismos made the language
used by classical authors such as Homer and Euripides the standard – an uneasy
mix of Ionian and Attic dialects, though Attic dominated because it was the
dialect of Athens. Other Greek dialects were often referred to, but everyday
koin? “common” Greek was mostly ignored. The tradition of favouring classical
literary language over everyday spoken language persisted through the Western
Classical Tradition into the twentieth century. The Greeks built up grammatical
analysis from nothing, and the framework which they passed on to the Latin
grammarians, if far from perfect, nevertheless supported a grammatical
tradition that has lasted 2,000 years. 13. The object-language is the language
under investigation. Apollonius and Priscian 111 Priscian In the period
following Apollonius, the Roman emperor Diocletian (fl. 284–305 CE) moved the administrative
capital of the empire east to Byzantium, which his successor Constantine (fl.
306–37 CE) renamed Constantinople. From the end of the fourth century,
Constantinople was the capital of the eastern (Orthodox) empire until it fell
to the Turks in 1453. With Constantine, Christianity became the official
religion of the Roman Empire, whose political, commercial, and cultural centre
from around 330 CE was Constantinople, while its spiritual centre remained in
Rome, where, by tradition, St Peter was martyred. Rome remained the capital of
the western empire centred on the Bishop of Rome (the Pope). From this time
onward scholars in the Western Classical Tradition tended to be Christian, with
a new ideological basis for philosophy, literature, and language. It was
believed that the original language must have been Hebrew, the first language
of God (Augustine 1984 XVI: 11). There were, however, three languages of God:
Hebrew, Greek (because the earliest Christian texts were written in Greek), and
Latin because St Jerome (Eusebius Sophronius Hieronymus, 340–42014) translated
the Old Testament from Hebrew and Aramaic into Vulgar Latin (contemporary
common koin? Latin); this became known as the ‘Vulgate’.15 Vulgar Latin shows
many features that were to emerge in Romance languages, but which are not found
in classical Latin. From earliest times Christian missionaries were involved in
language teaching to spread the gospel. St Jerome wrote a letter, Epistola 57,
on the theory of translation (Jerome 1963). Gothic (a Germanic language) is
known only through Ulfilas’ (c. 311–82) translation of the Bible into that
language, for which he reputedly created an alphabet based on Greek and Roman
alphabets (Wright 1954). St Cyril of Byzantium (c. 827–69) adapted the Greek alphabet
for Christianized Slavs thus creating the Cyrillic alphabet. As we shall see in
Chapter 7, in the early medieval period Christian missionaries gave impetus to
the teaching of Latin grammar. Since about 1934 the Summer Institute of
Linguistics (http://www.sil.org), in order to proselytize, has been developing
orthographies based on the Roman alphabet and IPA symbols and writing grammars
for the languages of many preliterate peoples. The study of classical Greek and
Latin literature required studying the grammar, which motivated linguistic
studies to concentrate on the language of the past; contemporary language was
ignored, and generally regarded as inferior to that of the classical authors.
Of the Latins, Varro was the most independent and original writer, but he was
peripheral to the main course of the Western Classical Tradition – perhaps
because his work was largely lost to Europe until the fourteenth century.
Quintilian is somewhat closer to the tradition, and the 14. Jerome was born in
Stridon in Dalmatia (inland from the Adriatic coast of Croatia),which was
sacked by the Goths in 377. Brought up a Christian, he studied classical Latin
with Donatus but eventually rejected the classical authors because of the pagan
content of their works. 15. St Isidore had a different explanation: ‘There are
three sacred languages, Hebrew, Greek, and Latin [… f]or it was in these three
languages that the charge against the Lord was written above the cross by
Pilate’ (Etymologies IX.i.3; Brehaut 1912: 208). See Chapter 7. 112 The Western
Classical Tradition in Linguistics
Latin grammarians that follow him (there are
more than we have space to consider16) stem more directly from the Alexandrian
tradition of Dionysius Thrax and Apollonius Dyscolus. However, all these scholars
have roots within the Stoic tradition.
As we have seen, the parts of speech
remained at eight, the article having been replaced by the interjection. The
relative pronoun, which had been called arthron by the Greeks, was relegated to
the pronoun class by Probus (first half of the fourth century CE) in Instituta
Artium (Probus 1961) and to the noun class by Priscian. Of the Latin
grammarians that endured, the most celebrated are Donatus (Chapter 5) and
Priscianus Caesariensis Mauri, born in Mauritanian Caesaria c. 490 CE, who
died, probably in Constantinople, c. 560. If he was Christian, he didn’t flaunt
the fact in his works. He was a prominent official in Constantinople under
Justinian I,17 where he wrote De Figuris Numerorum quos Antiquissimi Habent
Codices (on numerals, Keil (ed.) 1961, 3: 406–17), De Metris Fabularum Terentii
(on metrical feet in Terence, ibid. 418–29), Praeexercitamina (on aspects of
rhetoric, ibid. 430–40), De Accentibus (on pronunciation, ibid. 519–28) – as
well as describing and teaching classical Latin to people whose first language
was Greek. His work was extensive, of very high quality, and it has survived;
for those reasons Priscian is pre-eminent among the Latin grammarians. His
Institutiones Grammaticae “Institutes of Grammar” or “Grammatical Doctrine”
(also known as Institutionum Grammaticarum and here abbreviated to Inst.)
consists of 18 books constituting some 974 (admittedly heavily annotated) pages
in Keil (ed.) 1961, 3–4 (= Priscian 527). Most of the nearly 1,000 manuscripts
that survive are of Books I to XVI on sounds, word formation, and inflection;
these are known as Priscianus Major or De Octo Partibus, which was copied by
Flavius Theodosius, a clerk in the imperial secretariat of Justinian. Books
XVII to XVIII of Inst. (Priscianus Minor or De constructione) are on syntax.
Inst. was written in the years 526–27 while Priscian was teaching in
Constantinople; it is less obviously pedagogical than two later works:
Institutio de Nomine, Pronomine et Verbo “On the noun, pronoun and verb”
(Priscian 1961a, hereafter De Nomine) and Partitiones Duodecim Versuum Aeneidos
Principalium “Analysis of the first lines in the twelve books of the Aeneid”
(Priscian 1961b, hereafter Partitiones).
We will consider each of these in
turn.18 Priscian’s Institutiones Grammaticae
In the middle ages a manuscript
copy of Priscian’s Institutiones Grammaticae had a monetary value comparable
with owning a prestige car (such as a Rolls-Royce) today: in 1044 one was
exchanged in Barcelona for a house plus block of land (Law 2003: 128).
Book I
is on sounds and letters, and their part in constructing and differentiating
words, inflections and derivational morphs (to use today’s terminology).
Book
II is on the 16. E.g. Diomedes (Keil (ed.) 1961, 1) and Charisius (Keil (ed.)
1961, 1; Charisius 1964; Matthews 1994: 74–76). 17. Justinian was a Latin and
Greek speaking Illyrian (Illyrians were later swamped by Serbs, Croats, and
other Slavs). 18.
Unfortunately, there are no English translations of
Priscian’s major works.
Apollonius and Priscian 113 properties of syllables,
words, sentences, nouns and their accidence, proper nouns, common nouns,
adjectives, and their primitive and derived types.
Book III is on how to form
comparatives, superlatives, and diminutives. Book IV is on derivations from
nouns, verbs, participles, and adverbs and how to form them. Book V
distinguishes gender on the basis of word endings; then it discusses
grammatical number, compounds, and cases. Book VI examines nominative case as
expressed with vowel and consonant suffixes; then genitives in respect of both
final and penultimate syllables.
Book VII turns to the other oblique cases in
both singular and plural. Book VIII is on the verb and its accidence; Book IX
on regular conjugations; Book X on pasts and perfects. Book XI discusses
participles. Books XII and XIII are on pronouns. Book XIV is on prepositions;
Book XV on adverbs and interjections; Book XVI on conjunctions. Books XVII and
XVIII are devoted to de constructione sive ordinatione partium orationis inter
se “constructions or the arrangement among parts of speech” i.e. syntax.
Throughout Inst. Priscian makes very extensive use of literary examples from a
range of Greek and Latin authors; most of the last 100 pages of Book XVIII consist
of quotations from literature: literature was his corpus. The discussion of
sounds (voces) in Book I makes no advance on Donatus 1961a; however, the
discussion of letters includes quite a lot of Greek, including extensive
discussion of the Aeolic digamma ? and the old Greek aspiration marker? such
that ?? = ?, ?? = ? and ?? = ? (I: 24). Latin r is ‘without aspiration’,
whereas in word initial position and in geminates Greek ? is aspirated, cf. the
transliterated English words rhetor and Pyrrhus (I: 40). In I: 43 we find rules
like the noun and verb stem transforms in Tables 6.3 and 6.4 respectively.
Table 6.3. Graphemic simplification. N O M I N A T I V E GENITIVE –c apec+s ?
apex apicis –g +s greg+s ? grex gregis –ct noct+s ? nox noctis The third conjugation
verb stem with a present ending in –h changes to –x in the perfect, P: Table
6.4. A morphophonemic rule. h ? x veh– + P ? vex– trah– + P ? trax– Priscian
discusses syllable structure in Book II, but there is nothing new. ‘A syllable
is the sound of a letter uttered without pause in one pitch contour [accentus]
and one breath’ (II: 1). A Latin syllable can be from one to six letters long –
stirps is an example of the latter. Priscian includes phonotactic variations on
syllables, though he simply creates a list and proposes no phonotactic rule;
e.g. the consonant in the prefix ad- changes before c or g or p or t, e.g.
accumbo, accido, aggero, applico, appello, attingo, attineo; similarly for f –
affectus; l – allido; r – arrideo; n – annuo; and s – assiduus. (II: 7) 114 The
Western Classical Tradition in Linguistics The word is defined (II: 14) as the
smallest part of a sentence that makes sense as a whole. For instance, vires
“strength” cannot be divided into vi (“*”) + res (“thing”) because these do not
compose its meaning. ‘The sentence is a coherent sequence of words that makes
complete sense’ (II: 15). But Priscian does allow for one-word sentences as
responses, e.g. to the question “What’s best in life?” the answer “Honesty.” is
a good sentence, he says. A sentence is composed from a nominal signifying
substance or quality and an active or passive or deponent (vaguely described as
uter “other”) verb with the subcategories of mood, form (primitive or compound)
and tense, but not case. He distinguishes finite verbs from nonfinites and
participles. Pronouns are also sentence constituents. Adverbs are put with
verbs to complete their meaning (II: 20). A praepositio is one of two types:
prepositions that affect the case of a noun, e.g. de rege[ABL], apud amicum[ACC];
and prefixes that are affixed to words in compounds and do not affect case. For
example, in indoctus “unskilled” the prefix is negative and takes no case; but
the preposition in would take the accusative if illative (“into”) and the
ablative if inessive (“within”). The preposition inter normally takes the
accusative case, but in the compound verb intercurro “run between” none. The
preposition pro normally takes the ablative (pro bono publico “for the public
good”) but not in the compound noun proconsul. Finally, conjunctions like
either … or, both … and may link nouns and other case bearing words (such as
pronouns and participles) or verbs or adverbs. Some prepositions, but never
conjunctions, function like verbal prefixes, e.g. in subtraho, addico,
praepono, produco, dehortor. Priscian also distinguishes the adversative
connectives a(s)t and sed (“moreover, but”) without clearly identifying their
function. Next there is a section on nouns: A noun is the part of speech which
assigns a particular common or proper quality to persons or things. (II: 22)
Man and art are common nouns, one human, the other not; Virgil is a human
proper noun; the grammar of Aristarchus is a non-human proper noun – presumably
because it makes unique reference.
The primacy of nouns is not mentioned here,
but in Book XIV we find: ‘For the noun, which is the principal among all the
parts of speech …’ (XIV: 1); cf. Apollonius’ Synt. II: 4. Back in Book II we
are told that nouns have five subcategories (accidents): type (primitive or
derived), gender, number, shape (simple or compound), and case. There are
further extensive sections on patronymics and possessives.
Book III begins with
a long discussion of comparatives and the identifying morphology, then moves to
do the same for superlatives and diminutives. Book IV on denominals begins with
names derived from adjectives like Maximus and Catulus ? catulus “whelp”
diminutive of catus “clever, cunning”. Then there are derivations from verbs,
participles, other nouns, and adverbs, presented according to the form of the
denominal’s suffix. Patterns are presented (IV: 3–4) like amicus “friend.NOM”,
amici “friend.GEN”, amicitia “friendship.NOM”; Graecus “Greek.NOM”, Graeci
“Greek.GEN”, Graecia “Greece.NOM”; prudens “prudent.NOM”, prudenti
“prudent.DAT”, prudentia “prudence.NOM”. Although the regularity is shown,
there is no explanation offered for the root of the final denominals being a
dative; it is presented as arbitrary fact. Sometimes the seemingly obvious is
missed: Apollonius and Priscian 115 In ux: luceo, lux; duco, dux In nx et rx:
coniungo, coniunx; arceo, arx. (IV: 40) Both the second conjugation verb lucere
“to be bright” and the third conjugation ducere “to lead” (luceo and duco are
the first person singular present indicative) have the first person singular
perfect indicative luxi and duxi respectively, which offer a more convincing
source for the nouns “light” and “leader” than do luceo and duco. The roots are
probably luc- and duc- from which the nominal is derived by applying an –s
suffix; the digraph is simplified to , with which it is homophonous. The
explanation for the second pair is virtually identical. In fact, for the third
conjugation coniungere “to join together” it is identical, the root being
coniunc- (cf. the perfect participle coniunctum). The noun “spouse” is created
by adding –s ? *coniuncs ? coniunx. The perfect of the second conjugation verb
arcere “to hinder, keep out” is arcui or arc+vi; 19 its root is arc-, to which
–s is applied to create the noun arx “stronghold”. A discussion (in Book VI:
93f) of nominative and genitive case in nouns like vox~vocis, nex~necis,
fornix~fornicis, dux~ducis, grex~gregis suggests deleting the final –x of the
nominative and adding –cis; but we may instead take the –x as a simplification
of an underlying –cs, consisting of a nominative suffix –s on a stem ending in
–c; to the stem ending in –c, the genitive –is is suffixed. This is essentially
the rule exemplified in Table 6.3 (and I: 43).
Book V lays out the formal
correlations between final syllables and genders. There is not so much any new
insight here as the surveying of an immense amount of data. A section on number
begins with saying that the form of the word (he evidently means “noun”)
normally distinguishes the quantity being referred to. There is explicit
mention that we find only singular and plural because, unlike Greek, the mother
tongue of Priscian’s immediate audience, Latin has no dual, dualis apud Latinos
non invenitur (IV: 46). He pays some attention to effects of quantifiers on
nouns, verbs, participles and pronouns; and shows that adverbs are not
responsive to number (IV: 49f). Although there is sometimes ambiguity about
whether the form of a noun or pronoun is singular or plural (e.g. res, sui and
qui) this is never the situation with verbs (IV: 52). He discusses collective
nouns having singular form and plural meaning, e.g. populus “people, crowd”,
legio “legion”; and other nouns – all proper names – with plural form and
singular meaning, such as Athenae, Thebae (as in the Greek originals and also
English Athens, Thebes). Book V continues with discussion of the simple or
compound form of nouns, e.g. magnus, magnanimus, magnanimitas (“the great, the
magnanimous, magnanimity”) (V: 56). The book ends with a discussion of case.
Priscian begins with invariables (monoptota) such as letters and numbers:
hoc[NOM] alpha, huius[GEN] alpha; hi quattuor, horum quattuor. He proceeds
through nouns that have two, three, four, five, and six (hexaptota) case forms.
Book VI begins with a discussion of nominative and genitive cases. Priscian
says he is going to offer explanations and try to correct the errors of others
without malice (VI: 1). He looks at the masculine, feminine, neuter, and common
gender noun and pronoun suffixes.
Then he examines some variations in case
endings in different dialects or in older and 19. Letters and are
interchangeable in Latin. So too are and . 116 The Western Classical Tradition
in Linguistics contemporary Latin. Some genitives in –tis are on words derived
from Greek; he looks at nominals with short vowels, others with long vowels (or
long syllables). He identifies some female names from Greek that end in –on and
take a feminine article (i.e. demonstrative) but decline like a neuter noun in
–um e.g. ? ??????? haec Dorcium, ? ????????? haec Glycerium, ? ??????? haec
Chrysium, etc. (VI: 24). Later he demonstrates some masculine proper names from
Greek that behave otherwise – the pairs here are nominative and genitive:
?????? ????????, hic Memnon huius Memn?nis, ????? ???????, hic Sinon huius
Sin?nis, ??????? ??????????, hic Laocoon huius Laoco?ntis (IV: 29). No rule is
offered; readers are left to learn them by heart or invent their own rule. On
variations: Besides, Caesar declined “youth” pubis puberis; others, like
Probus, pubes puberis, and yet others puber puberis. (VI: 65) However, impubis
is the same in both nominative and genitive! Occasionally there is a rule
suggested: Masculine nouns ending in os make the genitive by deleting the s and
adding ris, e.g. lepos leporis, ros roris, mos moris, flos floris. An exception
is nepos nepotis “grandchild, nephew/ niece” which some people class as a
common noun despite nepotis being feminine.20 (VI: 68) One might compare this
with the common gender sacerdos~sacerdotis “priest(ess)”. And then Priscian
exemplifies some other nouns of common gender before looking at the case forms
of participles (VI: 77f). He points out that some trees denoted by second
declension feminine nouns look masculine; the demonstrative reveals their
gender: haec orn?s ~ huius orni “mountain ash”, haec fagus ~ huius fagi
“beech”, etc.
But quercus, pinus, ficus etc. (“oak, pine, fig”) are regular
fourth declension feminines with the genitive in –?s (VI: 83). Book VII, on
other cases, begins by listing the final letters of all nominals in Latin;
sometimes these are case morphs, but often Priscian omits the preceding
segments of the case morph; in other words, he does not here recognize case
morphs as such. However, once he starts discussing oblique cases of the first
declension (starting at VII: 3), he does indeed identify the morphs. He
identifies the first declension variations through the nominative and genitive,
then deals in turn with accusative, vocative, and ablative; first singulars,
then plurals. He draws attention to syncretistic or ambiguous forms, such as
the fact that the nominative and vocative plurals are formally identical with
the dative singular (VII: 9). Such homonyms can be disambiguated by co-text,
for instance, the third declension singular dative and ablative huic[DAT]
mari[“sea”] and ab hoc[ABL] mari 21 (VII: 55). Having finished with the first
declension he proceeds through the other four declensions. Book VIII is on
verbs. A verb is the part of speech with tenses and moods, without case,
signifying actions or being acted upon. […] Both intransitive (also called
‘complete’ [by the Stoics]) and deponent verbs, according to their nature, are
either active or passive. (VIII: 1) 20. There is a similar rule applying in haec
servit?s ~ huius servitutis, sal?s salutis, etc. (VI: 83). 21. The ablative
singular of this neuter noun can also be mare. Apollonius and Priscian 117 The
verb has eight subcategories: tense, mood, type, form, conjugation, person and
number, and meaning without reference to gender (VIII: 2). Deverbal nouns open
the discussion. A section on ‘signification’ (starting at VIII: 7) identifies
the link between an active (transitive) or intransitive form ending in –o with
a corresponding passive in –or. And here Priscian seems to be identifying
transitivity rather than voice. When I say ‘I hear you’, I say it because my
ears are being acted upon by your voice; and conversely, when I say ‘I am heard
by you’, my voice is making something happen in your ears. (VIII: 7)
Intransitive verbs (neutra) don’t have these properties. For instance one
cannot make passives out of spiro, vivo, ambulo, pergo (“I breathe, live, walk,
continue”) (VIII: 12f). These verbs are intrinsically active, but others are
intrinsically passive, e.g. rubeo, ferveo, caleo (“I am blushing, boiling,
warm”). Yet others express reciprocal (communia) acts that are both active (or
in fact deponent) and passive, e.g. osculor te “I kiss you” and osculor a te “I
am kissed by you”, criminor te “I accuse you” and criminor a te “I am accused
by you”. Deponents do not necessarily have this reciprocal property, cf.
conspicor te “I catch sight of you”, sequor te “I accompany you” (VIII: 14).
Later he identifies semantically related verbs such that one can say quaero a
te “I seek an answer from you” as an alternative to interrogo te “I interrogate
you” (VIII: 21). Then there are semantically and formally related pairs like
intransitive labo “I totter” and deponent labor “I fall down” (VIII: 36). The
section on verb tenses (de temporibus verborum) identifies three tenses, of
which the past is further divided into three: imperfect, perfect, and
pluperfect. The indicative mood freely takes all tenses. The imperative (VIII:
40f) accepts the present when we want something done immediately (quae statim
in praesenti volumus fieri sine aliqua dilatione) but is otherwise future. The
past imperative takes the form of the perfect or pluperfect subjunctive. The
optative is used for wishing and praying in all three tenses:
The optative,
however, although it also seems to pertain to the future – as we wish for those
things which we want to be given to us in the present or future – nevertheless
is used of the past, as well, since we often pray for what we lack or wish to
know of things we don’t know. (VIII: 42) The subjunctive occurs in all three
tenses. The tenses of the infinitive, participles and verbs in different voices
are then considered (VIII: 43–62). The moods indicate several different mental
attitudes. There are five of them: indicative or definitive, imperative,
optative, subjunctive, infinitive. (VIII: 63) Priscian fails to point out here
that there is no formal distinction made between the optative and the
subjunctive; they are semantic variants of the same clause-type (mood form).
Moreover, they interact in interesting ways with the imperative and infinitive.
Both points are mentioned in Book XVIII, where there is some discussion of the
relations among optative-subjunctives, imperatives, and infinitives; but the
lack of formal distinction between optative and subjunctive is dismissed.
Surprisingly, there is nothing in Priscian to compare with Apollonius’
treatment of mood, other than the likelihood that inclinationes animi (VIII:
63) is a Latinization of psuchich? diathesis. 118 The Western Classical
Tradition in Linguistics From a modern morphological standpoint the discussion
of verb types (species, primitive and derived) and forms (figura, simple and
compound) should probably be amalgamated under a common heading – even though
the discussion of types focuses on origin, that of forms on construction.
The
rest of Book VIII contains brief discussion of conjugation, person, and number
as verbal subcategories. Books IX and X present rules for declining verbs. For
instance, the past imperfect is formed from the present thus: Table 6.5.
Formation rule 1 for the past imperfect. In the 1st, 2nd, and 4th conjugation
ending in –eo, delete –s from the second person and add –bam: am?s ? am?bam
(Infinitive amare) doc?s ? doc?bam (Infinitive docere) ?s ? ?bam (Infinitive
ire) Table 6.6. Formation rule 2 for the past imperfect. For verbs of the 3rd
conjugation and those of the 4th in –io, the –o of the first person changes to
long –? and assumes –bam: lego ? leg?bam (Infinitive legere) facio ? faci?bam
(Infinitive facere) venio ? veni?bam (Infinitive venire) Matthews 1974: 72f
criticizes Priscian for deriving the past imperfect forms individually from a
specific present tense form. Priscian was probably assuming either (a) that his
students already knew Latin and he was teaching them how to relate grammatical
forms to one another; or alternatively (b) that the present tense forms would
be learned first and thus serve as a basis for new learning. The present tense
usually is learned before other tenses because it is the least marked tense and
can be realistically illustrated by using ostension in the teaching situation.
Thus what Priscian may have been doing is giving an effective enough rule to
teach the correct tense forms from a known model. To criticize his rule on the
basis of illogicality or psychological unreality in a language acquisition
situation, or for giving priority to one form rather than another when they are
all equal in the ear of the native speaker is to miss the point. Book XI is on
participles. Thus the participle is a part of speech which is used in place of
the verb from which it derives; it has gender and case like a noun and is
without verbal inflection for person and mood. (XI: 8) Uses for the participles
are demonstrated, e.g. ‘reading I learn for I read and I learn, and from my
teaching you learn for I teach and you learn’ (XI: 9). Book XII rehearses the
subcategories of pronouns. Book XIII opens with ‘Case is found on pronouns as
it is on nouns’ (XIII: 1). There are five cases except in second person
address, where there is a sixth, the vocative. Priscian extensively discusses
syncretism Apollonius and Priscian 119 among case forms. In Book XIII there is
a lot of comparison of Latin with Greek; for instance, Priscian contrasts
(formally) simple Latin adjectives with their compound Greek counterparts, and
vice versa: e.g. felix ??’?????, sanctus ??’????, pius ??’?????, et alia mille.
“fortunate, holy, pious, and a thousand others.” incestus ????????????, ineptus
?????, superfluus ????????, sicut et multa alia. “sinful, foolish, superfluous,
as well as many others.” (XIII: 17) Book XIV is on prepositions and prefixes;
and once again there is frequent reference to Greek. At the outset Priscian
follows in the footsteps of Apollonius:
And so it seems right to me that the
most learned Greeks, and above all Apollonius, whose authority in all things I
have acknowledged, should have proposed that the preposition is indeclinable;
therefore I begin from that. (XIV: 1) When discussing accents on prepositions
in XIV: 6 he claims similarity with Greek and in XIV: 9 he claims that where
Greek has only one preposition, Latin often has several: ut ???? pro circum et
circa et erga et de et super, quando memoriae est, ponitur; similiter ???? pro
apud et prope et praetor et propter. He notes that whereas in Greek a
preposition may cause the accent to shift on the stem, this is not so in Latin
except for the one adverb inde: ut ??????? ???????????, apud nos in uno
adverbio hoc solet facere, índe, déinde, éxinde, próinde, súbinde, … (XIV: 20)
He occasionally uses Greek in giving the meaning of a Latin word, e.g. De does
not only mean apo but also brings to mind peri as in ‘on [“concerning”] parts
of speech’; it is also a rendition of kata in some environments such as ‘bring
down’, ‘go down’, ‘take down’, ‘throw down’, ‘look down’, ‘look down upon’; it
is used for expectations in ‘seize’ … (XIV: 45) Rather curiously, but not
irrationally, Priscian counts the conjunctive suffix –que a preposition (XIV:
3). Book XV begins The adverb is an indeclinable part of speech which adds
meaning to the verb. The meaning of the adverb augments the meaning of the verb
in the same way that the adjective adds meaning to common nouns, e.g. a prudent
man acts prudently, a happy man lives happily. (XV: 1) Priscian points out that
the Greeks included interjections among the adverbs: ‘In hoe, au, heu, ei, oh,
ah terminantia interiectiones sunt, quas Graeci adverbiis connumerant’ (XV:
17); ‘Interiectum Graeci inter adverbia ponunt’ (XV: 40).
In Book XVI Priscian
identifies seventeen types of conjunctions.
The conjunct is an indeclinable
part of speech that connects other parts of speech with which it consignifies
showing augmentation or sequencing: there is augmentation when the same entity
is assigned different attributes as in Aeneas was both virtuous and brave;
there is sequencing when a consequence of things is shown, as in if he walks,
there is movement. Motion follows from walking, however walking does not always
follow from there being motion.
It is possible for things sitting and lying
down to move; but to walk without moving is not possible. (XVI: 1) 120 The
Western Classical Tradition in Linguistics That ends the Priscianus Major.
Priscianus Minor consists of the two very long Books XVII and XVIII on syntax.
In the preceding exposition of the parts of speech in this book we have mostly
followed the authority of Apollonius, not omitting input from others either
Latin or Greek where necessary; and we were even able to add something new
ourselves. In a similar manner we shall now speak of word order or syntax,
which the Greeks call suntaxis; we shall examine the data and if we should
discover patterns in the work of others or for ourselves, we shall not hesitate
to bring them forward. (XVII: 1) What Priscian says about the construction of
sentences is very similar to what we found in Apollonius’ Synt. I: 2. Letters
when they come together in an appropriate way form syllables and syllables form
words, so also words form sentences. (XVII: 2) To construct oratio perfecta
“well-formed speech” (XVII: 2), The noun is put in first place, the verb second
and to be sure no sentence without them is complete […] if you take away the
noun or the verb it makes the sentence incomplete. (XVII: 12) Shades of the
Stoics. Before the verb, therefore, it is necessary to put the noun; it is
proper for entities to act or undergo [“be acted upon”], in which is born the
position of the noun and from these the properties of the verb, that is active
and passive. (XVII: 14) The article merits a mention. Recall that Priscian’s
audience is Greek and he seems to be explaining that Latin speakers do not feel
the lack of a definite article. However Latin lacks a preposed article. When
declining the noun grammarians put the pronoun hic in front of it in place of
the definite article, as it is called. But there is never any consciousness of
the article in everyday speech. (XVII: 27) In XVII: 75f he points out the
incongruousness of Priscianus scribo “Priscian I.write” in comparison with
Priscianus scribit “Priscian writes[3.SG]”. Book XVII contains many analogia
like the following: tu meus filius legis, ego tuus pater doceo et tu Virgili
doces, ego Priscanius doceor “you, my son, learn [literally, “read”]; I, your
father, teach and you, Virgil, teach; I, Priscian, learn.” (XVII: 111) At least
once, Priscian the dry scholar reveals his human side: So indeed the genitive
of the possessives themselves is used in every case of secondary possession;
that is, whatever is possessed by a possession – as if a wife speaking of her
husband’s field or of some other thing [of his] should say iste ager mei est [“
that is the field of my [man]”], tou emou [“my masculine thing/person”], semine
mei praegnans sum [“I am pregnant with the seed of my [man]”], prolem mei
diligo [“I love the child of my [man]”], thalamis mei caste pareo [“I
virtuously obey my [man] in the bed-chamber”]. (XVII: 131f) ‘Mei’, translated
here as “my [man]” (perhaps it should be “my husband”), is the masculine
singular genitive of the first person pronoun. Priscian notes a variety of
transformational relations, though he does not consistently use any such term
for them. Apollonius and Priscian 121
The passive is made from the object NP
[per obliquos activo coniunctos verbo] being transformed into the appropriate
nominative form for the noun which is in concord with the passive verb; the
agent is transformed into an ablative as in ‘Caesar vanquished Pompey, Pompey
was vanquished by Caesar; I love you, you are loved by me.’ (XVII: 135) Other
examples of active to passive transforms are given in Book XVIII: 127f, and the
passive rule is also given, in similar terms to the above, in XVIII: 139. A
plural transform is exemplified: putting the singular [noun] with the singular
[verb] and the plural with the plural, when referring to one and the same
person intransitively, as in I Priscian write well and we orators write well.
(XVII: 153) Book XVIII is about the relationships between nouns and verbs
(XVIII: 1). For instance, We don’t say Priscianus lego[read.1.SG] nor
Apollonius legis[read.2.SG] [it is a solecism]; instead the way [to speak
correctly is] I, Priscian, read and You, Apollonius read. (XVIII: 4)22 In this
next passage Priscian recognizes omission of the relative pronoun. Similarly
with other cases following the nominative, either on a participle as discussed
above, or instead understood to necessarily belong to the nominative: my good
friend set off that is the one who is a good friend; a man with an honest face
[honest as to face] is perceived that is the one who has an honest face; the
swift footed man runs is understood as [the man] who is [swift footed]. (XVIII:
7) There is extensive discussion of all the cases and their different uses, for
instance from XVIII: 9 possessives using the genitive, dative, and ablative are
exemplified. Perhaps under the influence of Apollonius, Priscian recognizes an
underlying have in possessives, e.g. a man[NOM] of great virtue[GEN] for a man
having great virtue[ACC]; an unusually beautiful[ABL] woman[NOM] for [a woman]
having exceptional beauty[ACC]. (XVIII: 13) There are implicational relations
like me doctore[ABL ABSOLUTE], dum ego doceo “me being a teacher whilst I
teach” and virtute florente[ABL ABSOLUTE], quoad virtus floret “virtue is
flowering as long as virtue flourishes” (XVIII: 16). learned in grammar is a
participle […] student of grammar is a noun (XVIII: 21) He mentions such
lexical relations as the converses, father~son, master~servant, the contraries
friend~enemy, and some words similar in meaning like companion~ally,
near~neighbouring (XVIII: 26), but he does not demonstrate these relations
using constructions.
The closest Priscian comes to adopting Apollonius’
analysis of mood is just this one remark: From the infinitive does Apollonius
begin to explain regularities among verbs, showing it to be the generic form of
the verb; and for all verbs it is possible to express moods using it. (XVIII:
40) 22. ‘non dicimus Priscianus lego nec Apollonius legis; quomodo autem …
Priscianus ego lego et tu Apollonius legis [recte dicitur].’ 122 The Western
Classical Tradition in Linguistics He seems more interested that the infinitive
can occur with nominative, accusative and ablative nouns as in cognitus est
posse dicere, cognitum hunc posse dicere, cognito posse dicere, which the
English translation, roughly “what is known can be said”, does not
satisfactorily distinguish. cum dico enim bonum est legere nihil aliud
significo nisi bona est lectio. “when I say it is good to read I mean nothing
other than reading is good.” (XVIII: 43f) The contrast is between the
impersonal bonum[N.NOM] est with the infinitive and the nominal; bona[F.NOM] is
in agreement with the feminine noun lectio[F.NOM]. The discussion of moods is
disappointing after reading Apollonius. Priscian does identify and exemplify
the fact that one mood is used where we might, perhaps, expect another and he
would be on better ground if he were able to refer to the different
clause-types that can be used to express different illocutions. He recognizes
that imperatives are used to persons of equal and higher status as well as to
inferiors across a range of commands, exhortations, supplications, and
invocations (XVIII: 70f); and that As we show above, however, authors are found
to use the indicative, optative and subjunctive as an alternative to the
imperative. (XVIII: 74)23 Be aware also that imperative verbs joined with
vocatives make a well-formed sentence, as in [You] Apollonius teach, [You]
Trypho learn. [Their] nominative [counterparts] require the presence of verbs
[in the indicative] and at the same time require relative articles, cf. you who
are Apollonius teach, you who are Trypho learn. (XVIII: 75) Priscian mentions
in XVIII: 77 the fact that there is no formal distinction between the optative
and subjunctive in Latin but then ignores it, maintaining the contrastive
distinction inherited from Greek.
The optative is used in prayers and with the
adverb utinam “would that” (XVIII: 76); the subjunctive expresses doubt or
uncertainty (XVIII: 79). This subjunctive mood, therefore, examples of whose
force I have briefly gathered together, among the Latins is used when there is
doubt, approval, or possibility in cases where supposition is brought in.
(XVIII: 91) Priscian’s Institutiones Grammaticae is a massive work with the
grammatical points made being extensively illustrated from Greek and Latin
literature. It is poorly structured, a fact noted by early medieval
commentators, some of whom reorganized the material in excerpts so that, for
instance, all aspects of the noun were collected together instead of being
strewn across several books. One accessible example (because the Latin is
translated into English) is the Excerptiones de Prisciano (c. 980–90; Porter
2002), which was the source for Ælfric’s grammar of Old English (Ælfric 11th c.
manuscript; 1880). By comparison with Apollonius’ Peri Suntaxe?s, Priscian’s
Inst. lacks verve and originality. Priscian’s De Nomine Institutio de Nomine,
Pronomine et Verbo examines the inflecting word classes among the parts of
speech. It begins with the five noun declensions defined formally in terms of
the 23. See Allan 2006a: 46f for some discussion of this. Apollonius and
Priscian 123 genitive singular and exemplified using the demonstrative and
nominative as well. An explanation for the numbering of declensions is given in
the first sentence. All nouns used in good Latin are inflected according to
five declensions which are numbered in accordance with the sequence of vowels
forming their genitive case. The first, then, has its genitive ending in the
diphthong ae, e.g. hic po?ta [NOM “this poet”], huius po?tae [GEN “this
poet’s”]; the second has a long i as the case ending, e.g. hic doctus [NOM
“this learned man”], huius doct? [GEN]. The third [ends] in short is, e.g. hic
pater [NOM “this father”], huius patr?s [GEN]; the fourth in long us, as in hic
sen?tus [NOM “this senate”], huius sen?t?s [GEN]; the fifth in ei divided into
[two] syllables, e.g. hic mer?dies [NOM “this noon”], huius mer?di?i [GEN].”
(Priscian 1961a: 443) In this work Priscian is at least as thorough as
Apollonius. Even though Priscian lived in the eastern empire, where the
principal language was Greek, it is clear from this passage, as from all his
work, that his audience was assumed to be fluent in Latin. He was not teaching
Latin as a foreign language, although it is quite likely that the variety of
Latin he describes was not the Latin spoken by his contemporary audience.
He
quite often cites or quotes Greek authors and refers to the Greek language, for
instance, while exemplifying further the first declension nouns in Latin he
compares them with Greek nouns (ibid. 443): ut ?? ?????? ???˜ ?????? hic Lysias
huius Lysiae, ?? ’????? ???˜ ’???? hic Antas huius Antae, ?? ????????? ???˜
????????? hic Priamides huius Priamidae … Even if he didn’t expect his audience
to have a better knowledge of Greek than of Latin, he certainly thought that
they could usefully learn the grammar of Latin by comparing the two languages.
He introduces many additional comparisons for other declensions. He proceeds
through the various case forms in singular and plural for all declensions of
nouns before recommending (ibid. 449) further study of the topic in Books VI
and VII of Inst. With the same decisiveness we see in the opening discussion of
nouns, Priscian says that there are fifteen pronouns in Latin, divided into
eight ‘primitive’ pronouns (ego, tu, and six third persons sui, ille, ipse,
hic, iste, is) and seven ‘derived’ or possessive pronouns: meus, tuus, vester,
nostras, vestras, etc. are all first or second person; the third person
possessive uses a genitive of one of the third person primitives. The first and
second person primitives have only the one case form; the other cases are derivatives,
e.g. mei [GEN], mihi [DAT], me[ACC], a me [ABL]. Only the second person has a
vocative form and it is identical with the nominative: tu, vos. Next, Priscian
identifies nine ‘variable names’ (nomina mobilia, p. 449), which include
indefinite pronouns such as quis “who, what” and quantifiers like unus, nullus,
totus, alter (“one”, “none”, “all”, “other”) and discusses their inflections.
Then he turns to the four conjugations of verbs. Once again these are defined
formally. For example, Thus the first conjugation has the first person ending
in o which changes to long as to make the second person, e.g. amo [I.love] am?s
[you.SG.love]. The second conjugation has its first person in eo which changes
into long es in the second person, which is always in this conjugation a minor
syllable, e.g. doceo [I.teach] doc?s. (Priscian 1961a: 450)
Priscian reviews
tense and mood paradigms. Using the second person singular present indicative
suffix as a diagnostic he goes through all forms that share this characteristic.
For instance, the second person singular in –as: 124 The Western Classical
Tradition in Linguistics amas and all persons in the present indicative: amo
(1.SG), amas (2.SG), amat (3.SG), amamus (1.PL), amatis (2.PL), amant (3.PL)
amabas and all persons in the imperfect indicative: amabam, amabas, amabat,
amabamus, amabatis, amabant amaveras and all persons in the pluperfect:
amaveram, amaveras, amaverat, etc. legas and third and fourth conjugation
present subjunctives: legam, legas, legat, legamus, legatis, legant He proceeds
with all forms congruent with the second person singulars in es, e.g. second
conjugation present indicative doceo, doces, docet, docemus, docetis, docent;
first conjugation imperfect subjunctive amarem, amares, amaret, etc.; first conjugation
present subjunctive amem, ames, amet, etc.; third conjugation imperfect
subjunctive legerem, legeres, legeret, legeremus, legeretis, legerent. What
Priscian is doing here is demonstrating regularities that proceed from a known
source, namely, second person singular present indicative forms. Only after
dealing with the regularities does he turn to the few irregularities. He offers
transformational rules such as The pluperfect in all conjugations turns the
final i of the perfect into a short e and adds ram e.g. am?v? ? am?veram,
audiv? or audi? ? audiveram or audieram. (Priscian 1961a: 452) He discusses
imperatives, subjunctives (which, in this work, he always calls optativi
“optatives”), and infinitives before moving on the tense forms of participles and
supines.24 Here again he makes comparison with Greek. There is an intermission
where voice is exemplified and discussed.
To the first person in any
conjugation, tense or mood ending in o add r to create a passive, e.g. amo,
amor; amabo, amabor; legunto, leguntor (“I love, I am loved; I will love, I
will be loved; let them read, let them be read”). And there are comparable if
not such simple rules for constructing all other passives. He then returns to
present participles and deverbal nouns. And he ends up referring to his further
discussion of all these matters in Inst. Priscian’s Partitiones The Partitiones
Duodecim Versuum Aeneidos Principalium opens with the first line of Virgil’s
Aeneid followed by a discussion of the feet that appear in the verse and a more
general discussion of all the kinds of feet found in various types of verse:
dactyls – ? ?, spondees – –, anapaests ? ? –, iambs ? –, trochees – ?, etc.
Then follows a pattern of analysis repeated for the first lines of each of the
twelve books of the Aeneid. The verse is scanned, the caesuras identified and
the forms of the parts named.
The parts of speech are then discussed in great
detail along with their subcategories, their potential as the basis for
derivations, and other kinds of alternative forms.
Consider the discussion of
the first line of 24. Supines are a restricted set of verbal nouns, e.g. Hic
liber facilis est lectu[read.SUPINE] “This book is easy to read”; Romam
Caesarem visum[see.SUPINE] ivimus “We went to Rome to see Caesar”; mirabile
dictu[say.SUPINE] “amazing to relate”.
Apollonius and Priscian 125 Book V:
Interea medium Aeneas iam classe tenebat “Meanwhile Aeneas held the fleet in
the mid way.” The first paragraph reads: Scan the verse. Intere a medi um Aene
as iam classe te nebat. How many caesuras does it have? One. What? A
semiseptenarius, Interea medium Aeneas. How many are its elements (syllables)?
Ten; for it has three dactyls and two spondees. [The caesura, marked ||, occurs
in the preferred position after the long syllable at the beginning of the
fourth foot: ?nt?r?|? m?d?|[?m elided before a vowel] ?en?|?s || i?m | cl?ss?
t?n|?b?t. (Dactyl, dactyl, spondee, spondee, dactyl, trochee.)] Discuss each
foot. Interea etc. [this is skipped perhaps because the same kind of thing has
been done four times already].
How many parts of speech has this verse? Six.
How many nouns? Three. What? Medium, Aeneas, classe. How many verbs? One,
tenebat. What else is there? Two adverbs, interea and iam. Discuss each part of
speech. Interea is what part of speech? Adverb. What is an adverb? … (Priscian
1961b: 480) The style is reminiscent of the Ars Minor of Donatus (Donatus
1961b). Every part of speech is exhaustively discussed as it arises. There is
not much repetition in the twelve sections of this work because there is
something different to say about each word in the verse and its morphological
construction and particular type. E.g. interea “meanwhile” is a temporal
adverb; it is a derivation from the free (integro) element inter and the bound
(corrupto) part ea. The noun medium here is an excuse to compare (ibid. 481)
the definitions of nouns in Donatus (a part of speech with case signifying each
a thing or a proper name) and Apollonius (a part of speech signifying a
concrete or abstract thing with the quality of being either a common name or a
proper name).
Then all the properties of common nouns and adjectives are
examined. Proper nouns are examined in the discussion of the name Aeneas. Let’s
now skip to the end of the book where there is a parsing of the proper noun
Latinos from the first line of Book XII of the Aeneid. Latinos is what part of
speech? Noun. What sort? Derived, masculine gender, singular number, simple
form. Of what type? Derived. Nonetheless it can be the proper name of a thing,
the father-in-law of Aeneas, a Latin person from Latium, or possessive in
Latinus ager “Latin field” [. Also Latium is the name of something concealed or
broad. Latinitas [“Pure Latin”] is [derived] from Latin [the language]; from
Latium also derives Latius [“something from the region”] and Latiaris
[“Jupiter, patron of the Latins”]. Make a compound noun. Latinigena [“Latin
people”] in the manner of Troiugena [“Trojans”], Graiugena [people of Greece].
But in the latter the second of the geminate i’s changes to u making Troiigena
and Graiigena Troiugena and Graiugena [respectively]. (Priscian 1961b: 515)
This sufficiently demonstrates Priscian’s method in Partitiones and explains
why it was much used in teaching Latin. Anyone who associates the word parse with
studying Latin grammar has this work of Priscian’s to thank. A summary of
Priscian’s contribution Priscian describes the word as a minimum unit of
sentence structure. A sentence is an arrangement of words expressing a complete
thought.
The eight parts of speech are largely semantically defined and indeed
Priscian says – following Apollonius:
‘the parts of speech cannot be correctly
known without knowing their proper signification’ (Priscian 527 II: 17).
Priscian’s definition of the noun made no reference to it taking case
inflection (Priscian 527 126 The Western Classical Tradition in Linguistics II:
22), although he did deal with the various subcategories of the noun, just like
Donatus had done. This suggests a presumption that the case inflections were
already known to his audience. The verb was given the same description as in
the Techn? Grammatik? and the Artes of Donatus. Noun.
Assigns a common or
particular quality to every body or thing. Verb. The part of speech with times
and moods signifying action or being acted upon. Participle. A part of speech
deriving from the verb and having at the same time gender and case. Pronoun. Is
substituted for proper nouns and can take specific persons. (This is
descriptively inadequate.) Following Apollonius, Priscian 527, XIII: 31 says
that the pronoun indicates substance without further properties. Adverb. A part
of speech without inflection (i.e. it is indeclinable) which is added to the
meaning of the verb.
Preposition/prefix.
Is indeclinable and put before other
parts of speech in either apposition or composition. [...] It is put before and
in apposition to case-inflected words, and in composition with both cased and
non-cased words. Conjunction.
An indeclinable part of speech conjoining other
parts of speech which are to be understood together, and showing the
relationship between them. Interjection. A class of words syntactically
independent of verbs, indicating feeling or state of mind. Priscian talked
about the properties or essence of each part of speech, e.g.
Proprium est
nominis/verbi/pronominis/adverbii, etc. and it is clear that he thought the
semantic content of the parts of speech was their most significant property.
But he also discussed their subcategories, thus giving some account of their
morphological properties. However, semantic and functional descriptions
dominated the Western Classical Tradition. Priscian’s morphological description
of nouns and verbs took the relatively unmarked nominative and present
indicative forms from which are derived the other forms by a process of letter
changes. Varro’s important insight into the difference between derivational and
inflectional morphology was ignored by Priscian, who followed the Alexandrian
practice. Case was shown to rest not on forms of any one noun or declension of
nouns but on semantic and syntactic functions systematically correlated with
differences in morphological shape over the system as a whole. The many-to-one
relations between form and meaning (use) are clearly allowed for in his
analysis. In looking at case forms he uses the concord with various established
case forms of the demonstrative hic/haec/hoc “this.M/t.F/t.N” as normative
models. In analysing the verb he used the same model as is found in the Techn?
Grammatik?, which was not entirely applicable to the Latin verb. He did
recognize that the perfect and the aorist are syncretic in Latin.
However,
unlike Varro, he did not recognize aspect.
Like Donatus, he recognized a
difference between optative (expressing wishes) and subjunctive (other
hypotheticals) in Latin, following the Greek verb morphology, although this is
not superficially distinct in Latin verb morphology (as it is not in English).
Here we have an Apollonius and Priscian 127 early example of a formally
nonexistent category being carried over from one language to another. It is
very much like talking about the nonexistent (surface) cases of English nouns
as many traditional grammars did. Priscian held that the normal constituent
order in the Latin sentence of subject (nominative noun or pronoun) before verb
is natural because substance is prior to the action it performs or undergoes:
Before the verb, therefore, it is necessary to put the noun; it is proper for
entities to act or undergo, in which is born the position of the noun and from
these the properties of the verb, that is, active and passive. There is the
idea of a subject noun in the verb itself, which without the noun cannot be
meaningful. (XVII: 14) This idea stretches back to Aristotle, if not Plato; it
is certainly found in the Stoics. There are of course many languages in which
the subject noun phrase does not precede the verb; but none of these would have
been known to grammarians before the nineteenth century. It is foolhardy to
make statements about naturalness on the basis of one or two related languages.
The terms subject and object were not in use in Priscian’s time, although
reference was occasionally made to a logical subject or topic.
Priscian did
notice that transitive verbs govern an oblique case.
There is not much
systematic analysis of Latin sentence structures in Priscian’s opera, but there
is something on relations between parts of speech and sentence structure in so
far as nouns and verbs are recognized as primary, and other parts of speech as
subordinate to them.
Thus there would be a sort of governance hierarchy: *
sentence governs nouns and verbs; * nouns govern prepositions and pronouns; *
verbs govern adverbs. Subordination was only acknowledged when one whole clause
is subordinated to another clause. Conjunction was not recognized as either
coordinating or subordinating.
In his detailed application of the Greek
grammatical framework to Latin, Priscian achieved the goals of Roman
grammarians from the previous five hundred years.
Inst. is less consistently
reliant on formal criteria than its Greek predecessors, but Priscian makes
greater use of semantic criteria.
In fact, no other Greek or Latin grammarian
appears to have achieved the same degree of preciseness in formal descriptions
of parts of speech as are found in the extant version of the Techn? Grammatik?.
Priscian stands at the end of late antiquity.
The next five or six hundred
years are the early middle ages. Up to here, the chapters in this book have
been sequenced chronologically as well as thematically. Chapters in the rest of
the book are thematic with sequenced internal chronology, but the chronology
within the chapters overlaps with the chronology within other chapters. This
allows for a more coherent structure than a rigid chronological matrix would.
Subscribe to:
Post Comments (Atom)
Posts
No comments:
Post a Comment