BY JULIUS GOEBEL, PH. D., NEW YORK CITY. The question concerning the origin and nature of the so-called impersonals is one of great interest to the philosopher as well as to the philologian; for it is here where logic and grammar meet to solve a problem that belongs to both of these sciences. The impersonals and the sentences formed by them, seem to present an exception, if not a contradiction, not only to the syntactical rule that every sentence should consist of a subject and a predicate, but also to the law of logic, according to which a judgment is composed of two members. A linguistic investigation of this problem can, therefore, not be made without seeking aid from logic and psychology, and the results of such an inquiry may help to throw light on the relation of grammar to logic, and on the psychological basis of language. We could, of course, also proceed exclusively according to the historical method by inquiring after the origin of the impersonals in general as well as in particular cases. But such an investigation, too, would have had to be assisted by psychological and logical considerations without which it would scarcely rise above a mere collection of material. Before we attempt, however, an explanation of this apparently irrational linguistic phenomenon we may do well to cast a glance at the history of the various attempts made in the direction of a solution of the problem.

Considering the fact that the principal question concerning the impersonals has always been whether the sentences formed by them contain a subject or not, we can distinguish two schools of writers on this question: those who affirm the existence of a subject and those who deny it. The supporters of the theory that a subject is contained in the impersonals, may again be divided into different classes, according as they claim that an indefinite or a definite notion is the subject of such sentences. We are assured by many grammarians that the indefinite notion forming the subject is contained in the verbal stem of the impersonal. Thus the verb-form curritur is to be explained : cursus curritur, an explanation which appears as far back as the grammarian PRISCIAN who says: Cum dico curritur, cursus intelligo, et sedetur sessio, et evenit eventus. It is obvious, however, that this explanation is not sufficient for all impersonals and when I say gespeist, there can not be any doubt as to what gespeist wird, das Speisen, the verbal stem, or die Speise, the food.

The indefinite notion may also be supplied by other means as some grammarians think. In ver we can add Zeus, in συσκοτά ζαι we can supply ἡμέρα and instead of es fehlt an Geld we can say Geld fehlt. But it would be very difficult to find a subject for every impersonal, and this lack of uniformity has especially displeased the philosophers who tried to find an indefinite notion which would cover all cases. Many of them believe to have found it in the indefinite Etwas, which they substitute for the German es. Thus UEBERWEG says: Niemals kann einem Urtheil und Satze das Subject völlig fehlen, wohl aber kann die bestimmte Subjectsvorstellung fehlen und anstat dessen das blosse Etwas (es) eintreten. In es ist ein Gott, es gibt einen Gott, tritt die unbestimmt vorgestellte Totalität des Seienden oder ein unbestimmter Theil desselben als Subject ein, gleichwie auch in den Sätzen: es regnet, etc. Similar views are represented by Lotze who says: Wer ein impersonales Urtheil ausspricht, betrachtet den bestimmten Inhalt als haftend an einem unbestimmten Subject. Das Es in 'es blitzt' bezeichnet den allumfassenden Gedanken der Wirklichkeit, die bald so, bald anders gestaltet ist.-BRANTL, BERGMANN, WUNDT and STEINTHAL may also be named here as representatives of the same view, which we must regard as a product of philosophic speculation, but which has no support in the facts of common speech and thought.

Entirely antagonistic to this view is the opinion which holds that there is no subject contained in the impersonals. Dissatisfied with the attempts to find a subject, as mentioned above, a number of logicians and grammarians regard the impersonal as an exception in human speech, some even going so far as to require a revision of the laws of logic concerning the nature of

the judgment. Among the philosophers we find HERBART and TRENDELENBURG expressing doubt as to whether the impersonal really contains a judgment. Thus the former declares: dass das Impersonale nicht als gewöhnliches Urtheil anzusehen sei, and TRENDELENBURG regards it als ein unvollständiges Urtheil, ein Rudiment eines Urtheils.-Unconcerned about the logical consequences, grammarians like Heyse, Grimm and BENFEY have declared that the impersonals do not possess a subject. They were recently joined by MIKLOSICH whose book, 'Subjectlose Sätze,' caused the whole discussion to be opened again with renewed energy. His position was defended by A. MARTY in an essay: "Ueber subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie" in the Zeitschrift für wissenschaftliche Philosophie, viii, 56 ff. The results which MIKLOSICH and MARTY apparently had reached by their investigations were, however, attacked and refuted by W. SCHUPPE in an essay on "Subjectlose Sätze" in the Zeitschrift für Völkerpsychologie and in the most interesting and instructive monograph by SIGWART: "Die Impersonalien, Eine logische Untersuchung."

A peculiar position between the two factions is occupied by PAUL in his 'Principien der Sprachgeschichte.' Making a distinction between the psychological and the logical subject of a sentence, he ascribes the former to the impersonals while he denies the existence of the latter. We shall see, however, in the course of this paper which will attempt to supplement SIGWART'S logical discussion by linguistic considerations, to the effect that a complicated question like this can not be settled by a few remarks of an apodeictical nature, as PAUL seems to think. In view of the greater abundance of impersonals in the German language than in other European idioms, I shall take my illustrations mostly from this source.

The attempts at an explanation such as is mentioned above, are characterized by two modes of procedure, one of which we, too, might be tempted to adopt. It might be possible that the impersonals which have been preserved to us in the various languages, are the relics of a pre-historic stage of syntactical construction from which the present sentence has gradually developed. This is, however, shown to be a fruitless hypothesis by the simple fact that the impersonals are constantly increasing, not to speak of the fact that we are entirely without means to prove such a development in the languages whose history we are able to trace. Having, therefore, to deal not with a prehistoric relic, but with a living linguistic phenomenon, still capable in the German, at least, of new formations, we might also reduce the various impersonals to their common characteristics and construct upon these an hypothesis which would explain them. Since all impersonals agree in containing either a verb ; for example, ΰει, tonat, es blitzt 'it lightens', or a noun, or an adjective with the verb to be; for example, es ist kalt, 'it is cold', is it not possible to find a subject which would meet all cases though it is not expressed?

It will be remembered that the philosophers, especially, have tried the latter method without, however, reaching satisfactory results. For it has been justly remarked by SCHUPPE and SIGWART that a conclusion from that which has been expressed to that which was thought, is not admissible, since many things which are thought do not, or cannot find an expression in language. The want of an expressed subject in the impersonals is, therefore, not to be considered a proof of either its existence or its non-existence. But since we are constantly making use of impersonals we are still in the position of observing the psychological process at the basis of this peculiar syntactical construction. An investigation into this psychological process may not only solve the problem concerning the real nature of the impersonal, but it may also furnish us with a key to the understanding of its origin in the various languages. There is no doubt that we meet in the impersonal the same psychic function which appears at the basis of all human speech, still it may not be useless to repeat here a few fundamental facts. We know that the picture of the world about us consists of representations caused by sense-impressions, representations which find their expression in the various words of the language. But in order that we may use a word to signify a certain representation, it is necessary that its meaning be known to the speaker as well as to the hearer; it is necessary that by association the word shall have become the sign of the special representation. If I use the word "man," I imply that by previous sense impressions the picture of the human form is present to my mind and that this inner, mental picture is closely connected with the word "man." Thus, the words of a language present a vast number of representations which I can reproduce in others that speak the same language, by the simple utterance of the words. But while I may thus reproduce, in the hearer, representations which he already possesses, I am unable to impart any new knowledge by such isolated words. The latter can only be accomplished if, to the word used, a relation to some other representation is added which is originally not contained in that word. It does not make any difference whether this relation is expressed by the speaker or not; for, if it is not, it has to be supplemented by the hearer. When I pronounce the single word "march!," it will produce in your minds the representation of marching which you have previously gained by looking at soldiers or freemasons. If, however, I were to address a company of soldiers with the word "march!," the term would have still another meaning besides the representation of the idea of marching. The soldiers would supplement the notion with the concept that I also meant the command of marching. If, however, the hearer is not able to supplement by the situation the relation existing in my mind, it is necessary to express it. Hence it is evident that human speech, as far as it is a statement or the conveying of a judgment, must necessarily consist of at least two members which grammar distinguishes as subject and predicate.

While we have thus deduced the fact that the logical judgment contained in every sentence implies the existence of a subject and a predicate, we must still determine to what class of judgments the impersonals belong in order to determine their subjects. This inquiry is facilitated by the character itself of the impersonals which, as is generally conceded, express, with a very few exceptions, concrete perceptions. SIGWART in his 'Logik' has fortunately named the judgments expressing a synthesis of a perception made at this moment with the representation that I already have in my mind: Benennungsurtheile. Such a judgment I express, for example, when I show a friend the city in which I live and which he has never seen before, and, pointing out to him the various buildings, I say: this is the cathedral, that is the court house, etc. A similar synthesis, though the mental process is somewhat more complicated, is contained in judgments expressing a quality or an action of some object; as, for example, the apple is green, or the bird flies. This process of synthesis takes place very easily where we have to deal with phenomena, or objects, which are accessible to our observation. It becomes, however, more difficult when the object, or the phenomenon, is removed from my direct observation and I perceive only the quality, or the action.