II.

Types of Sentence Structure in Latin Prose Writers.

BY PROF. CLARENCE LINTON MEADER,

UNIVERSITY OF MICHIGAN.

THE traditional method of studying Latin sentence structure from the point of view of style the method elaborated with considerable detail in Nägelsbach's Stilistik in the chapter entitled "Architektonik der Rede" — is, at least in some respects, inadequate. This will be apparent to any one who follows carefully the attempt of Schmalz in his Lateinische Stilistik, pp. 465 ff. (3d ed.) to give even a sketchy outline of the history of Latin sentence structure. It is the purpose of this paper to direct attention to a different method of dealing with the subject, one which may perhaps not so much supersede as supplement the somewhat mechanical system of counting clauses and participles and examining the order in which they follow each other, or the manner in which they are interwoven or interlocked. "Le stil c'est de l'homme même." We can more fully understand an author's style after we have determined what mental processes were involved in the organization 1 of his sentences. We must transfer our attention from the outer product to the inner process which it represents. The traditional method is admirable as applied to a rhetorical writer like Cicero, but fails when applied to a spontaneous writer like Tacitus. Cicero may be studied from the point of view of rhetoric; Tacitus should be approached from that of psychology.

This paper has taken as its basis the system of psychology elaborated by Professor Wilhelm Wundt of Leipzig. It is therefore necessary to describe briefly the two processes (or rather groups of processes) which he terms association and apperception. Both consist in the uniting (Verbindung) of

1 Gliederung in the terminology of Wilhelm Wundt. As this word implies, the present paper accepts the definition of a sentence given in his Völkerpsychologie, I, vol. ii, ch. 7, § I, 5, c. (All references to this work apply to both editions.) See also Wundt, Sprachgeschichte und Sprachphilosophie, pp. 68-71, and Jl. of Germ. Phil. IV (1902), p. 390.

2 For a detailed discussion of these processes see his Physiologische Psychologie, passim, and his brief but lucid Grundriss der Psychologie, 5th ed. pp. 243–334.

psychical elements or concepts (Begriffe) into ideas (Vorstellungen). The main difference between them lies in the greater or less part which the feelings play in them, as also in the form which the feeling assumes. Speaking relatively the associations are passive processes; we give ourselves over to the flow of ideas, and they form and run as if of their own accord. A revery is a series of such associative acts; dreams afford a still more striking example. Into the apperceptive processes the feelings enter more largely, and in particular, they assume that most complex form, the will. Accordingly, in the apperceptive processes we exercise a more or less strong control over our ideas. There is a feeling of activity (Tätigkeitsgefühl). The will enters in, and we relate, compare, analyse, and synthesize the elements of our ideas, whereas in association the logical relations between the concepts are less regarded. The distinction between the two is at bottom merely relative. Some forms of association are more active than others; some forms of apperception are more passive than others; and there are various links (Uebergangsstufen) between them. Into most of our mental activity both processes enter, but one or the other may greatly predominate. It is characteristic of apperceptive analysis that it always proceeds by a series of bisections (dichotomy), the unit of thought (Gesammtvorstellung) that forms the basis of the sentence being first divided into halves, each of which is again subdivided, and so on.

How are these processes reflected in language? The written or spoken sentence is, of course, simply the outer form corresponding to a series of associative and apperceptive processes. The basis of the sentence is a relatively complete unit of thought of varying magnitude and complexity, which is present in consciousness at the moment the organization of the sentence begins. By successive acts of analysis and synthesis the various elements of this unit are set into their logical relations to each other. Each single act of apperceptive analysis yields only two sub-units (dichotomy), e.g. subject and predicate, substantive and attribute. The connection is therefore a closed one (in the sense that no third

1

member is possible) and is symbolically represented in Wundt by a curved line, thus: a b, a and b representing the sub-units. A purely apperceptive sentence will contain only connections of this kind. Under favorable mental conditions, however, any one of the elements of a sentence may induce an associative addition to the original unit, and this addition (or the original element) may induce a second addition and so on indefinitely. The connection between these associative additions and the inducing elements is therefore an open one. The formula a ̄a1a aα will then represent a series of such associative processes, a being the original element, and a1, a, a, a, being induced elements. The associative (open) nature of the connection is symbolized by the horizontal line. I know of no more striking illustrations of these two forms of connection than the Sanskrit copulative (associative) and determinative (apperceptive) compounds respectively. The formula just given represents the compound devagandharvamănuṣo (=a—u) raga-rākṣakās, "gods, heavenly singers, men, serpents, demons" (to correspond to the last horizontal stroke in the formula, an adi "and so forth" might have been added). The determinative compounds, on the contrary, always admit of dichotomic analysis; e.g. devadutaḥ, messenger of the gods," is a closed connection: ab. In general, subordinate conjunctions will mark the apperceptive connections, coördinate the associative. In primitive thinking (eg. in that of children) and imaginative compositions (e.g. in poetry) associations predominate, in scientific thinking the apperceptive processes rule.

Dr. Boucke, of the University of Michigan, has entered upon an historical examination of the German and English literatures from this point of view and has formulated a number of types of sentence structure.1 These are as follows: 2

1 See Jl. of Germ. Phil. IV (1902), pp. 389-420.

2 The symbol U means " unit of thought." All other letters represent subject

or predicate. ab AB symbolizes an ascending sentence (subordinate clauses

ab

first), symbolizes a descending sentence (main clause first).

id

1) isolating [a) simple, b) ascending, c) descending]:

5) synthetical [a) ascending, b) descending]:

1

For a description of these types we must refer to Dr. Boucke's admirable paper, and limit ourselves to a few very general statements. Types I, 2) and I, 3) take their essential characteristics from the two main types of imagination, the passive and the active (or intuitive and combinating) respectively. In II, 4) and II, 5) are reflected the two more complex apperceptive processes, analysis and synthesis, while II, 2) and II, 3) show an apperceptive groundwork with an intermingling of associative elements, II, 3) being the more complex. The differences between I, 1), II, 1), and II, 5) lie not simply in the organization of the unit of thought itself, but perhaps even more in the relations between successive units, the first showing open connections, the last closed connections, the other relative disconnectedness or isolation. A very good concrete notion of literary style that is based upon the predominating associative form of thought (the primitive type) is given by the following passage from Reuter's Stromtid, cited by Dr. Boucke (p. 395): "So Hawermann sat there and his hands were folded and his honest blue eyes turned upward and a more beautiful light was mirrored in them than that of God's sun. Then a little maiden came running up and laid some daisies in his lap and his prayerfully uplifted hands sank and were thrown round the child: it was his child—and he rose up from the bench and took his child on his arm and in his hand he had the flowers and went with his child along the path down the garden." Examples of the isolating type

will be cited below.

On undertaking the study of the Latin sentence, we meet at once several conditions which make the problem somewhat different from that of the student of modern languages. In the first place it appears to be quite generally accepted as true that the ancient languages, particularly Latin, make extensive use of connecting particles which bring out clearly and distinctly the relations existing between the successive units of thought, while the modern languages, the French in particular, leave those relations unexpressed.1 If this is true, we should expect to find types I, 1) II, 1), and 1 It is possible that this reputed difference is merely formal.