Problems in Reconstructing Schopenhauer’s Theory of Meaning: With Reference to His Influence on Wittgenstein

Abstract

The article contributes to the discussion of Schopenhauer’s possible anticipation of both the representational theory of language and the use theory of meaning and the reception of his philosophy by early and late Wittgenstein. Schopenhauer’s theory of language is presented and brought into the context of these two theories. His use of the terms “word,” “concept,” and “meaning” is analyzed and it is shown that he applies them ambivalently. The article’s main findings include a demonstration of how Schopenhauer’s ambivalent terminology enables the twofold interpretation of meaning: as representation-based and use-based.

1 Introduction

With the following article I seek to make a contribution to the discussion regarding Arthur Schopenhauer’s theory of language and especially his possible anticipation of questions later elaborated upon by Ludwig Wittgenstein. Given the fact that the latter is known to have been a reader of the former, this discussion also includes questions about Schopenhauer’s influence on Wittgenstein.

In my article, I refer to Jens Lemanski’s article Schopenhauers Gebrauchstheorie der Bedeutung und das Kontextprinzip (Schopenhauer’s Use Theory of Meaning and the Context Principle) in the Schopenhauer-Jahrbuch of 2016 (cf. [5]). In it, Lemanski argues that there are significant parallels between Schopenhauer’s theory of language as presented in his Philosophical Lectures [13] and the one presented by late Wittgenstein in Philosophical Investigations [17], even though comparative research until now has focused mainly on similarities between Schopenhauer’s The World as Will and Representation [11] and early Wittgenstein’s Tractatus Logico-Philosophicus ¹ [16].

I consider it crucial for the further investigation of the question to point out some problems in Schopenhauer’s theory of language that have not yet garnered much attention. To do so, in Sect. 2, I start with a concise reconstruction of the main features of Schopenhauer’s theory of language, based on which I show its parallels with the representational and use theories. Then, in Sect. 3, I point out some fundamental problems with Schopenhauer’s formulation of his theory of language. I do this in reference to what I first formulated in my article from 2015 (cf. [2]) and then elaborated more extensively in Chapter 1 of my book Begriff und Methode bei Arthur Schopenhauer (Concept and Method in Arthur Schopenhauer’s Philosophy) (cf. [1])—neither of which is available in English. I specifically point out the ambivalent use of the terms “concept” (“Begriff”) and “word” (“Wort”) in Schopenhauer’s writings. In Sect. 4 I proceed to demonstrate how this problem influences the interpretation of which theory of meaning Schopenhauer actually supports. I show that both the representational theory and use theory can be founded upon this ambivalence. In Sect. 5 I analyze some additional issues that appear in connection with his use of the term “meaning”—namely, his anticipation of the separation of sense and reference. In Sect. 6 I show that even though two theories of meaning might be ascribed to him, he strongly favors one over another. In the last section I provide a brief summary of the findings and propose some future research considerations.

2 Schopenhauer’s Theory of Language and its Links to the Representational and Use Theories

Even though Schopenhauer’s theory of language has been discussed by a few scholars, unlike some topics appearing in his philosophy, it has never been the main focus of Schopenhauerian research. Apart from one article by Rudolf Malter in which the most crucial elements are discussed (cf. [7]), there is little literature dedicated solely to the problem of language. A significant motivation for investigating Schopenhauer’s views on language seems to have come from the analysis of his impact on Wittgenstein and the analytical tradition in philosophy. However, these analyses were comparative and the need for a holistic approach to Schopenhauer’s philosophy of language has remained unaddressed by researchers. For example, Lemanski points out that Wittgenstein scholars mainly refer to para. 9 of The World as Will and Representation when discussing Schopenhauer’s views on language [5, p. 183]. However, this paragraph is not sufficient for a definite account of what Schopenhauer had to say on this matter. A planned new German edition of Schopenhauer’s Philosophical Lectures (Philosophische Vorlesungen), a work which contains several long passages referring to logic, language, and concepts, may stimulate new research.²

To understand the position of the language problem within Schopenhauer’s philosophy, it has to be pointed out that in his philosophical investigation of what the world is he assumes the stance that our experience is made up of two different epistemological dimensions: (1) what we experience by our senses and (2) what we experience by our minds.³

We find this division of human experience into two classes in para 9. of The World as Will and Representation. The first class, which is referred to as “the real external world,” is made up of “representations of perception” (“anschauliche Vorstellung”) [11, p. 39]. The second class, which is referred to as “reflection,” consists of “non-perceptual” or “abstract” representations (“nichtanschauliche Vorstellungen”) [11, p. 40]. Notice that Schopenhauer refers to these two classes of experience using the same term “representations,” and differentiates them into (1) perceptual and (2) non-perceptual. The reason for this is his adoption of the Kantian point of view wherein all experience is phenomenal as it is constructed by the subject. In Schopenhauer’s terminology, what we experience as the real external world and as our thoughts are, from the metaphysical perspective, different modes of how we represent the world to ourselves. Thus both the external and abstract dimensions of the world can be understood as forms of representation.

In this context it should also be mentioned that Schopenhauer’s system includes the metaphysics of will by which the world as representation is reduced to a mere appearance of the will. From this perspective it is difficult to talk about a real external world. However, as a starting point for the presentation of his philosophy, Schopenhauer uses the perspective of the world as representation, not as will. In the first book of volume one of The World as Will and Representation he assumes the perspective that the “world is my representation” [11, p. 3]. It is significant for the reconstruction of his theory of language that para. 9, which is dedicated solely to the problem of concepts and can be considered the most coherent presentation of his theory of language from works published during his life, is also included in this book and within this perspective. Therefore, this seems to be the main perspective from which he analyzes the problem of language.

As has been said, Schopenhauer identifies the class of representations of perception with what we might call the real external world, i.e., with the world of sensory objects. On the other hand, the second class of representations, which are non-perceptual, is identified with “concepts.” These “form a peculiar class existing only in the mind of man, and differing entirely from the representations of perception”⁴ [11, p. 39]. This means that the two classes of representation differ strongly from each other. However, Schopenhauer also remarks that abstract representations “stand in a necessary relation” to the representations of perception, without which “they would be nothing” [11, p. 40]. This stance is based on the assumption that “the abstract representation has its whole nature simply and solely in its relation to another representation” [11, pp. 40–41]. Such a representation might also be an abstract representation or a representation of perception. Additionally, Schopenhauer makes a remark that strictly determines the relation of reflection and perception: “the whole world of reflection rests on the world of perception as its ground of knowledge” [11, p. 41].

What this means for the relation between abstract and perceptual representations, or, accordingly, the relation between conceptual and sensory knowledge, is summarized in the second volume of The World as Will and Representation:

It has been shown that concepts borrow their material from knowledge of perception, and that therefore the whole structure of our world of thought rests on the world of perceptions. It must therefore be possible for us to go back from every concept, even if through intermediate stages, to the perceptions from which it has itself been directly drawn, or from which have been drawn the concepts of which it is in turn an abstraction [12, p. 71].

Succinctly, Schopenhauer seems to state here that to understand the concepts we have in our minds we need to refer them to what we consider to be objects of the real world. He thus gives a clear epistemological priority to knowledge gained from the senses and subordinates conceptual knowledge to experience.

This stance strongly resembles Wittgenstein’s formulation of what Lemanski refers to as the representational theory of language (cf. [5, p. 173]). Its manifesto can be found in the Tractatus Logico-Philosophicus: “3.203. The name means the object. The object is its meaning […]” [16]. Another formulation, the one used by Lemanski in his article, appears in the first paragraph of the Philosophical Investigations. There, Wittgenstein argues with this theory and formulates it with the following words: “the individual words in language name objects (‘Gegenstände’)” [17]. In both formulations this theory assumes that elements of language (names resp. words) represent real objects.

Schopenhauer’s understanding of concepts as presented above appears to share the same thought with the representational theory—this is even reinforced by the fact that he refers to concepts as “representations of representations” [11, p. 40]. However, it has to be remarked that there is a significant difference between Wittgenstein and Schopenhauer’s account of the relationship between reality and language. Schopenhauer makes the additional distinction between words and concepts. It can be found throughout his main philosophical writings and is well explained in a passage from On the Fourfold Root of Sufficient Reason:

Now as representations, thus sublimated and analysed to form abstract conceptions, have, as we have said, forfeited all perceptibility, they would entirely escape our consciousness, and be of no avail to it for the thinking processes to which they are destined, were they not fixed and retained in our senses by arbitrary signs. These signs are words. In as far as they constitute the contents of dictionaries and therefore of language, words always designate general representations, conceptions, never perceptible objects; […] [14, p. 116].

Based on this quotation, the difference between words and concepts seems quite obvious: words are sensory symbols of concepts and concepts are mental objects that are represented by words. A very similar account of their difference can also be found in Schopenhauer’s Lectures, where he says that a word is “the sensory sign of the concept” (“das Wort: es ist das sinnliche Zeichen des Begriffs”)⁵ [13, p. 243]. This claim in fact states that there is an ontological difference between words and concepts: the first belong to the dimension of sensory reality (i.e., perceptual representation), whereas the second belong to the dimension of reflection (i.e., non-perceptual representation).

Thus, there is at least one significant difference between the representational theory of language as presented by Wittgenstein and Schopenhauer’s stance. It can be summed up in the following way.

• Wittgenstein: individual words in language name objects.

• Schopenhauer: individual words in language name concepts, which represent objects.

The crucial difference between these two stances is that in Wittgenstein’s formulation of the representational theory words are considered representations of real objects, whereas in Schopenhauer’s theory words are not representations of real objects, but of mental objects which he calls “concepts.” These mental objects are representations of real objects.

This difference brings up the problem of the carrier of meaning. In Wittgenstein’s account the meaning of a word is explicitly identified with an object: “In this picture of language we find the roots of the following idea: Every word has a meaning. This meaning is correlated with the word. It is the object for which the word stands” [17, para. 1]. In Schopenhauer’s theory we have to ask the question whether it is a word or a concept that has meaning. If it is the word, then its meaning will be understood as the thought of a specific person, a mental object called “concept.” If, on the other hand, concepts are found to be the carriers of meaning, then the meaning will be the real object that is represented by a concept in the mind of a person.

Therefore, at this stage of our investigation, in order to answer the question of what kind of language theory Schopenhauer formulates it should suffice to determine whether he considers words (i.e., sensory signs of language) or concepts (i.e., mental representations of real objects) to be the carriers of meaning. If concepts are the carriers of meaning, we could say that his theory of language is similar to the representational theory in Wittgenstein.

However, if words are the carriers of meaning, we would have to underline that it does not accord with this theory. Such a view would instead resemble what is called the use theory of meaning. Its formulation, according to Lemanski (cf. [5, p. 172]), can also be found in late Wittgenstein’s Philosophical Investigations. Arguing against his own earlier view on language, Wittgenstein proposes a new theory of meaning, as an alternative to the representational one: “the meaning of a word is its use in the language” [17, para. 43]. In this account, the meaning of a word is not identified with objects, i.e., elements of reality, but with its use in different contexts. Or, we might perhaps say, with how people use it. If we consider the possibility that, for Schopenhauer, the word is the carrier of meaning, then meaning will have to be understood as the mental state of a subject using it. This already bears some similarity with late Wittgenstein’s account. In both cases meaning is not an element of reality but something in the minds of people using language.⁶ Thus, if we were able to determine that Schopenhauer proposes that words are the carriers of meaning, we would find substantial evidence for his influence on late Wittgenstein’s understanding of language.

However, it is difficult to determine whether Schopenhauer holds that words or concepts are the carriers of meaning, because, contrary to his claims that words and concepts differ ontologically and that he strictly distinguishes them, he often uses the terms “word” and “concept” as names for what seems to be the same designate.

3 The Ambivalent Use of the Terms “Concept” and “Word”

Schopenhauer’s claims on the difference between words and concepts have encouraged scholars to treat this distinction as an important and indisputable element of his theory of language. Lemanski, although being somewhat careful, says that in almost all of his texts Schopenhauer strictly distinguishes between the terms “word” and “concept” (cf. [5, p. 187]). This is understandable. In a number of passages Schopenhauer indeed reassures us that he clearly distinguishes these terms as shown in the section above (cf. [14, p. 116], [13, p. 243]). However, we cannot rely on this distinction when discussing his theory of meaning. His sharp theoretical distinction between words and concepts is not reflected in his use of the terms when writing about the problems of language. A number of passages can be found in which he breaks away from self-imposed definitions of both words and concepts. Significantly, this happens when he dedicates his attention to examining what concepts actually are and how they function.

A hint that the distinction between words and concepts is in fact rather blurred can be found in some articles on Schopenhauer’s theory of language. For example, Gerhard Mollowitz obviously struggles to keep them apart when he writes about the “difficult border between abstract creation of words and concepts” (“schwierige Grenzpunkt der abstrakten Wort- und Begriffsbildung”). He later coins the term “conceptual words” (“Begriffs-Worte”) to adequately describe Schopenhauer’s theory of how thoughts are expressed in language [8, p. 53]. If the difference between words and concepts were as clear as claimed by Schopenhauer, there would be no such problem at all.

To find out the reason for this problem it is necessary to enhance his claims with an analysis of how he actually uses these two terms. Above all, he clearly states that words and concepts are connected to each other so strongly that we usually fail to notice the difference between them:

Thus when we read or listen, we receive mere words, but from these we pass over to the concepts denoted by them so immediately, that it is as if we received the concepts immediately; for we are in no way conscious of the transition to them [12, p. 23].

In other words, the claim that there is an ontological difference between words (sensory objects) and concepts (mental objects) is relativized here by the observation that from the perspective of the subject it is difficult to tell them apart. It is also expressed in the following remark from the Lectures which states that we are actually unable to grasp our concepts without words that represent them: “I should give you example of a concept for which there is no word: this is impossible” (“denn ich soll Ihnen als Beispiel einen Begriff mittheilen, für den es kein Wort giebt: das geht nicht”) [13, p. 244]. Thus, it seems that the claimed ontological difference between words and concepts is enhanced here by the claim that there is almost no epistemological difference between them. Or, in other words, their difference cannot be experienced by the subject.

This is further supported by the position that a person cannot have intellectual access to concepts (which are, by the above definition, mental objects of some kind), unless they are represented by words:

In this property they [i.e. concepts] have, to a certain extent, an objective existence that yet does not belong to any time-series. Therefore, to enter the immediate present of an individual consciousness, and consequently to be capable of insertion into a time-series, they must be to a certain extent brought down again to the nature of particular things, individualized, and thus linked to a representation of the senses; this is the word. Accordingly, this is the sensible sign of the concept, and as such is the necessary means of fixing it, in other words, of presenting it vividly to the consciousness that is tied to the form of time, and thus of establishing a connexion between our faculty of reason, whose objects are merely general universalia knowing neither place nor time, and consciousness which is tied to time, sensuous, and to this extent merely animal [12, p. 66].

Yet another interesting feature of concepts is also contained in this argument. They are presented as kinds of “objective” entities, i.e., something outside of the subject’s consciousness. If it were not for the words that fix them in consciousness, we would be unable to grasp them. If we stick to the definition of words as sensory signs which Schopenhauer gives in the above passages from the Fourfold Root and the Lectures, we might conclude that conceptual thinking is possible mainly when we are speaking aloud or writing, i.e., using sensory signs. However, here he is apparently speaking of mental processes when he refers to using words for bringing concepts into the subject’s consciousness. Taking the matter strictly, this is inconsistent with his claim that words are sensory. There are other passages where he does this. For example, here he directly states that we can also have words in our thoughts:

Of course, it sometimes happens that concepts occupy consciousness even without their signs, since occasionally we run through a chain of reasoning so rapidly that we could not have thought of the words [nicht hätten die Worte denken können] in so short a time. But such cases are exceptions that assume great exercise of the faculty of reason, which it could have attained only by means of language [12, p. 66].

What Schopenhauer means here is that sometimes, when reasoning very quickly, our thinking does not include the use of words. However, this should be treated as an exception. He seems to assume that the thinking process usually consists in thinking words, which obviously is not a sensory but a mental process. This opens up a whole new dimension of interpretation of what can be understood by the term “word.” Words are not merely sensory signs but also something like mental signs. They are used for “fixing” concepts. Not only does this modify Schopenhauer’s initial claim about the sensory ontology of words, but it also brings up new questions about what the term “concept” refers to. Concepts can no longer be defined simply as the mental objects that words refer to. The ontological differentia specifica between words and concepts described in the preceding section (i.e., their sensory or mental nature) cannot be upheld. Now both words and concepts have to be understood as kinds of mental objects, as elements of our mental processes. Given this, a new way of differentiating between these two types of mental objects needs to be found.

Solutions can be found in several passages of Schopenhauer’s works where he indicates that concepts are something which we are not aware of as long as they have not been represented by words. This can already be seen in the quotation above where he claims that concepts have a certain kind of “objective existence” (cf. [12, p. 66]). In the Lectures he similarly states: “Usually we become aware of a concept only together with its sign, the word.” (“In der Regel werden wir uns des Begriffs immer nur mit seinem Zeichen, dem Wort, zugleich bewusst”) [13, p. 244]. This means that words and concepts understood as mental objects differ in the following way: words are within the scope of our consciousness and concepts are, at least in most cases, outside of it, since we are usually not aware of them. In this picture conceptual thinking becomes something that occurs without our willful participation and, in order to become aware of our thoughts, we need to represent them with symbols in our minds, to which Schopenhauer refers using the term “words.”

Therefore, we can distinguish at least three types of objects with different ontologies that Schopenhauer indicates when reflecting upon the nature of language. For the sake of clarity I am going to use non-Schopenhauerian terminology to name them and from now on I will refer to them by the terms in parentheses. These objects are as follows:

• (i) sensory signs that represent mental objects (sensory signs),

• (ii) mental signs within the scope of our consciousness (mental signs),

• (iii) mental states outside the scope of our consciousness (notions).

Having made this distinction, we are able to formulate the main problem with Schopenhauer’s use of the terms “word” and “concept”: he uses these two terms to refer to the three types of entities listed above.

The ambivalent use of the term “word” has already been shown above. At one time Schopenhauer claims that words are sensory signs, and later speaks of words as mental signs. Thus, let us now investigate the ambivalent use of the term “concept.”

In the Lectures Schopenhauer reflects upon a possible situation: struggling to find the proper word to express a concept. From this he concludes that concepts might be grasped without words, however, we need words if we want to grasp them on demand:

However necessary words are for thinking and however much a concept needs a sign; the necessity of a sign does not consist in the fact that without it the concept could not be grasped at all, that it could not be thought of (because it can in and for itself, as we often lack a word for expressing a concept we have). It rather consists in the fact that the willful [willkürlich], arbitrary evocation of the concept is possible only through the sign: the sign does not serve for thinking it, but for making it present at any time. Thus, it would be false to argue with the necessity of the signs for concepts for the assumption that during thinking and speaking we actually solely operate with signs and they completely represent [vertreten] concepts; […]

(So nothwendig auch zum Denken die Worte sind und so sehr auch der Begriff eines Zeichens bedarf; so beruht dennoch die Nothwendigkeit des Zeichens nicht darauf daß ohne dasselbe der Begriff überhaupt gar nicht gefaßt, gar nicht gedacht werden könnte (denn das kann er an und für sich, da oft uns ein Wort fehlt unsern Begriff auszudrücken), sondern darauf, daß die willkürliche, beliebige Hervorrufung des Begriffs nur durch das Zeichen möglich ist: das Zeichen dient nicht ihn zu denken, sondern ihn jederzeit zu vergegenwärtigen. Darum wäre es falsch wenn man aus der Nothwendigkeit der Zeichen für die Begriffe die Annahme begründen wollte, daß wir beim Denken und Reden eigentlich ganz allein mit den Zeichen operirten, und sie völlig die Begriffe vertreten; […]) [13, p. 247].

Here “concepts” seems to refer to entities of type (iii) which might perhaps be described by the English term “notions”—i.e., cognitions the subject has but cannot grasp with full clarity on demand. This understanding of concepts also appears on page 244 of the Lectures where Schopenhauer clearly states that, in the case of abstract concepts such as “justice” or “power,” it is impossible for the subject to fully grasp them if they are not mediated by specific words (cf. [13]). The general idea that thinking happens with notions that are not fully evident to the subject itself unless they are represented by mental signs probably finds its fullest expression in the metaphor of the surface of water:

To make the matter clear, let us compare our consciousness to a sheet of water of some depth. Then the distinctly conscious ideas [Gedanken] are merely the surface; on the other hand, the mass of the water is the indistinct, the feelings, the after-sensation of perceptions and intuitions and what is experienced in general, mingled with the disposition of our own will that is the kernel of our inner nature. Now this mass of the whole consciousness is more or less, in proportion to intellectual liveliness, in constant motion, and the clear pictures of the imagination, or the distinct, conscious ideas [Gedanken] expressed in words, and the resolves of the will are what comes to the surface in consequence of this motion. The whole process of our thinking and resolving seldom lies on the surface, that is to say, seldom consists in a concatenation of clearly conceived judgements; although we aspire to this, in order to be able to give an account of it to ourselves and others. But usually the rumination of material from outside, by which it is recast into ideas [Gedanken], takes place in the obscure depths of the mind. This rumination goes on almost as unconsciously as the conversion of nourishment into the humours and substance of the body. Hence it is that we are often unable to give any account of the origin of our deepest thoughts; they are the offspring of our mysterious inner being [12, pp. 135–136].

This quotation gives a picture of how the process of thinking functions according to Schopenhauer. It can, however, also be seen in the context of Schopenhauer’s theory of language by referring to his definition of “thinking”: “What is properly called thinking, in its narrowest sense, is the occupation of the intellect with conceptions (‘Begriffen’)” [14, p. 119]. In other words, for Schopenhauer, thinking is being intellectually concerned with concepts. Consequently, the passage above might be interpreted as a metaphor for how concepts are dealt with by our minds. It shows that Schopenhauer assumes the existence of some kind of conceptual thinking that is not fully within the scope of the consciousness of the subject, from which we might conclude that concepts are to be understood as mental states (notions) outside of our consciousness.

In the quotations above it therefore seems that Schopenhauer is presenting a theory wherein conceptual thinking is a process that takes place independently from language. Words are in turn elements of language necessary for giving an account, to others and ourselves, of the effects of conceptual thinking. Using the chosen metaphor, we might say that words and language are on the water’s surface, whereas concepts are in its depths. Interestingly enough, I have come across only one scholar who actually realizes this problem. Only Jankowitz describes concepts in Schopenhauer’s theory as “extralingual reality categories” (“übersprachliche Wirklichkeitskategorien”) [4, p. 65].

However, this understanding of concepts is not consistently represented by Schopenhauer. This can be seen in the following passage:

In other words, it must be possible for us to verify the concept with perceptions that stand to abstractions in the relation of examples. Therefore these perceptions furnish us with the real content of all our thinking, and wherever they are missing we have had in our heads not concepts, but mere words [12, p. 71].

Here “concepts” are juxtaposed against “mere words,” with their differentia specifica being the reference to perception. Schopenhauer seems to assume that “mere words” and “concepts” are experienced by the subject in the same way - as mental objects so similar to each other that we might not be able to distinguish them from one another at all. He seems to be speaking of mental objects (the ontology of type ii above) that are either “mere words” or “concepts,” depending on whether they have or do not have reference in the real world. The same way of understanding concepts can be found in several other passages. For example, when Schopenhauer criticizes the use of very abstract concepts he is obviously referring to elements of language that the subject is conscious of:

Every philosophy which […] takes as its starting-point arbitrarily chosen abstract concepts such as, for example, the absolute, absolute substance, God, infinite, finite, absolute identity, being, essence, and so on, floats in air without any support, and so can never lead to a real result. However, philosophers have at all times attempted it with such material; [12, pp. 82–83].

According to the picture contained in the water metaphor, a subject has no influence on what happens in the depths of its unconsciousness, i.e., what notions are created there. Thus, Schopenhauer’s intention here is obviously to criticize the use of elements of language that refer to very general concepts. If by “concepts” he meant unconscious notions, his criticism would become nonsensical, as it would refer to something the criticized author had no impact upon.

Another example is found where Schopenhauer says that “the most special concept is almost the individual and thus almost real; and the most universal concept, e.g., Being (the infinitive of the copula) is scarcely anything but a word” [12, p. 64]. Again, by using the terms “word” and “concept” he seems to be referring to signs of language and he treats both of them as ontologically equal.

In summary, it seems that Schopenhauer uses the terms “word” and “concept” to refer to the three types of entities listed above in the following way:

• By the term “word” he sometimes refers to entities of type (i), i.e., sensory signs, and sometimes to entities of type (ii), i.e., mental signs,

• By the term “concept” he sometimes refers to entities of type (ii), i.e., mental signs, and sometimes to entities of type (iii), i.e., notions.

In other words, by “words” Schopenhauer might mean both sensory and mental signs for the subject’s cognition. By “concepts” he might sometimes mean mental signs for such cognition, and sometimes the content of cognition, i.e., a notion that cannot be grasped clearly and voluntarily by the subject without a sign. Additionally, he differentiates the entities of type (ii) into mental signs with or without semantics, which he calls “mere words” and “concepts,” respectively. This means that he might refer to entities of type (ii) as “words,” and elsewhere talk of ontologically similar entities as “concepts,” when he will intend to stress that signs with semantics are meant. And yet, elsewhere he will underline that they are completely different: “Yet the concept is entirely different not only from the word to which it is tied, but also from the perceptions from which it originates” [12, p. 63].

All this brings up a serious problem in discussing Schopenhauer’s theory of meaning. His own terminology lacks a clear distinction between words and concepts. He sometimes uses these terms as synonyms, when referring to mental signs with semantics, but sometimes as antonyms, when distinguishing between signs and their semantic content.

Consequently, within the framework of Schopenhauer’s own terminology, it is almost impossible to answer the question whether it is words or concepts that are the carriers of meaning, at least without a precise analysis of the specific context of each passage in which the terms “words” and “concepts” are used. Even though Schopenhauer introduces an apparently strict ontological difference between words and concepts (i.e., sensory and mental objects) it is lost in his further elaborations where he also seems to distinguish signs of language (both sensory and mental) and the extralingual notions they refer to using the same terminology. Significantly, none of these uses can be considered final.

4 Consequences of Ambivalence for Understanding Schopenhauer’s Theory of Meaning

In Sect. 2 I showed that the classification of Schopenhauer’s theory of language depends on what we determine to be the carrier of meaning in his theory of language: words or concepts. In Sect. 3 I demonstrated that the definitions of words and concepts are not strict. Consequently, settling on a definition of meaning within his theory becomes difficult.

These difficulties are iterated in Lemanski’s above-mentioned article from 2016. Referring to Schopenhauer’s Lectures, Lemanski points out that what Schopenhauer means when using the expression the “real value of words” (“wahren Werth der Wörter”) is precisely the same as the “meaning” of the word, i.e., the concept signified by the word, which makes up the semantics of the word [5, p. 187]. On the other hand, a few sentences later, Lemanski clearly states that the distinction between words and concepts lies in the fact that concepts have meanings, whereas words are sensory symbols of concepts, just like digits are symbols of numbers [5, p. 187]. This happens again when Lemanski discusses the theory of foreign language acquisition and states that Schopenhauer describes this process as representing a semantically empty word from a foreign language with a meaningful concept of one’s own language [5, pp. 187–188]. From this it might be concluded that, on the one hand, a concept makes up the semantics of a word (i.e., a word can have or not have meaning depending on whether it represents a concept or not) and, on the other hand, it is also possible for concepts to either have or not have meaning (when he is speaking of a meaningful concept, it seems that it is also possible for a concept to be meaningless). Of course the question also has to be answered whether the semantics of a word, i.e., a concept, is the same as its meaning. As has been shown in Sect. 3, Schopenhauer sometimes speaks of meaningless words in distinction to meaningful concepts. By “words” and “concepts” he would then mean the same type of mental objects present in the consciousness of the subject. In other words, in his terminology, a meaningless word is just an “empty word,” whereas a meaningful word is called a “concept” (cf. [12, p. 71]).

Let us sum up the claims that result from reading Lemanski’s account of Schopenhauer’s theory of meaning:

• (C1) words have meanings which consist of the concepts they represent,

• (C2) concepts have meanings.

Claim (C1) can be found directly in Schopenhauer’s quotation from the Lectures which Lemanski refers to:

[…] only from the different context in which a word is found do we abstract its true meaning, do we find the concept which the word describes.

([…] erst aus dem verschiednen Zusammenhang in dem man das Wort findet abstrahirt man sich dessen wahre Bedeutung, findet den Begriff aus, den das Wort bezeichnet) [13, p. 246].

Claim (C2) is made by Lemanski based on the argument provided by Schopenhauer on p. 243 of the Lectures where he says that words are merely sensory symbols of concepts. Significantly, Schopenhauer himself does not clearly state that concepts have meaning (in fact he does not even use the term “meaning”/“Bedeutung” on this page).

This does not mean that Lemanski is misinterpreting Schopenhauer’s intention. In other quotations Schopenhauer speaks directly about concepts, not words, as entities having meaning. For example: “I wish to trace back to their proper meaning these concepts of good and bad” [11, p. 395]. However, in this quotation, by the term “concept” he seems to be referring to a mental sign whose semantics he is about to examine. As shown above, in other passages he is referring to mental signs with semantics as “words.” Putting it simply, Schopenhauer’s intention here is to examine the true value of the sign “good,” i.e., he wants to analyze the notion which the word “good” signifies. Several additional passages can be found where Schopenhauer discusses the meaning of concepts in a similar sense, i.e., he expresses his intention to analyze the semantics of a sign of language. In the following example the meaning of a concept is made explicit:

It is true that, so far as the abstract representation, the concept, is concerned, we also obtained a knowledge of it according to its content [Gehalt], in so far as it has all content [Gehalt] and meaning [Bedeutung] only through its relation to the representation of perception, without which it would be worthless and empty [11, p. 95].

Here he refers to the semantics of a concept by a number of terms that seem to be synonymous: “content” (“Gehalt”), “meaning” (“Bedeutung”), and “relation to the representation of perception” (“Beziehung auf die anschauliche Vorstellung”). He also indicates that concepts might have value (as they might also be “worthless”) and content (as they might be “empty”). The last expression is a clear reference to his axiomatic stance that concepts are “representations of representations,” i.e., they are abstract/mental representations of other representations and eventually can be traced back to representations of perception. It therefore seems that, here, Schopenhauer is presenting the stance that concepts have meaning that consists of their connection to perception.

Based on the above selections, we might therefore distinguish two different theories of meaning formulated by Schopenhauer by enhancing the previous list:

• (T1) words have meanings which consist of the concepts they refer to,

• (T2) concepts have meanings which consist of the sensory objects they refer to.

These two theories of meaning are based on two fundamental assumptions about language made by Schopenhauer:

• (A1) signs of language (either sensory or mental) signify mental states (notions),

• (A2) mental states (notions) draw their content from the perceptions that are the foundation for them.

This is the core of the problem with determining which theory of meaning Schopenhauer is actually formulating. By the term “meaning” he sometimes means mental states (notions), and he sometimes means perceptions (elements of sensory reality).

Additionally, none of these theories of meaning refers strictly to what has been described here as notions, i.e., concepts understood as extralingual elements of unconscious mental processes. Rather, both theories result from the fact that Schopenhauer tends to interchangeably use the terms “words” and “concepts” when referring to signs of language. Simultaneously, the second theory is apparently formulated in reference to Schopenhauer’s investigation of the nature of concepts understood as notions.

These two theories are connected with what was described in Sect. 2 as the representational and use theories of language. The first one anticipates the use theory: signs of language signify mental states, which are private, and therefore the only way to find out their meaning will be through the analysis of how they are used by the speaker. The second one anticipates the representational theory: all our notions are founded in perception and their meaning can be drawn only in reference to real objects.

It thus seems that depending on whether the term “words” or “concepts” is discussed, Schopenhauer presents one theory or another. In Sect. 6 I am going to show that he also puts them into a hierarchy. However, before this is done, the question of the meaning of concepts requires further investigation.

5 Meaning as Sense and Reference

In the preceding sections I tried to show that Schopenhauer uses the term “meaning” in reference to two different things: mental states, when he is talking about the meaning of the signs of language, and empirical objects, when he is talking about the meaning of notions. This ambivalent use of the term is connected with his blurred and inconsequential distinction between words and concepts.

However, it has to be underlined that in most cases Schopenhauer seems to avoid the term “meaning” when he speaks of “concepts.” Instead, as shown in the quotation in the section above (cf. [11, p. 95]), he often uses the German term “Gehalt,” translated by Payne as “content.” When he gets more specific, he clearly distinguishes between the “extent” (“Umfang”) and “content” (“Inhalt”) of a concept, a distinction that is lost in the English translation in which both the German terms “Gehalt” and “Inhalt” are translated as “content.” This can be seen in the following passage which we are somewhat familiar with from above:

Further, since the content [Inhalt] and extent [Umfang] of concepts are in inverse relation to each other, and thus the more that is thought under a concept, the less is thought in it, concepts form a sequence, a hierarchy, from the most special to the most universal, at the lower end of which scholastic realism, and at the upper end nominalism, are almost right. For the most special concept is almost the individual and thus almost real; and the most universal concept, e.g., Being (the infinitive of the copula) is scarcely anything but a word [12, p. 64].

He also elaborates on this issue in the Lectures when discussing the use of spherical diagrams for representing concepts.⁷ Here we gain further insight into what is meant by these two terms:

The relative size of the spheres refers consequently not to the size of the content [Inhalt] of the concepts, but to the size of the extent [Umfang]: not the concept in which we think the most (the most qualities) has the broadest sphere, that is not the concept richest in thoughts; but the one through which we think most things: that is the one which is the quality of many things.

(Die verhältnißmäßige Größe der Sphären bezieht sich also nicht auf die Größe des Inhalts der Begriffe, sondern auf die Größe des Umfangs: nicht der Begriff, in welchem wir das meiste (die meisten Eigenschaften) denken, hat die weitere Sphäre, also nicht der gedankenreichste Begriff; sondern der durch den wir die meisten Dinge denken: also der welcher eine Eigenschaft sehr vieler Dinge ist) [13, p. 271].

From these two passages we can see that Schopenhauer clearly distinguishes between the two following semantic capacities of a concept: (1) the capacity to refer to specific qualities and (2) the capacity to refer to empirical objects. He also says that these two capacities are indirectly proportional to each other, i.e., the more specific qualities a concept refers to, the fewer empirical objects it refers to. Or, conversely, the higher the number of empirical objects the concept can refer to, the lower the number of specific qualities of the objects it can refer to. This is strongly connected with Schopenhauer’s stated claim that the most abstract concepts are the “emptiest and poorest,” [14, p. 116] whereas “all truth and all wisdom ultimately lie in perception” [12, p. 74].

This distinction has much in common with Gottlob Frege’s distinction of “sense” and “reference.” Frege, talking about a sign defined as “name, combination of words, letter,” understands its reference as “that to which the sign refers,” and its sense as “the mode of [its] presentation.” He then gives the famous example: “The reference of ‘evening star’ would be the same as that of ‘morning star,’ but not the sense” [3, p. 57]. In other words, he expresses the thought that although “evening star” and “morning star” refer to the same object, the use of each of these terms brings up different associations for the subject. Schopenhauer seems to understand the content (“Inhalt”) of a concept as a bundle of features of an object that the subject has grasped and associates with the object. Frege says that the “sense of a proper name is grasped by everybody who is sufficiently familiar with the language or totality of designations to which it belongs; […]” [3, pp. 57–58]. The similarity to Schopenhauer consists of the realization that even though we might think that a sign of language refers to specific objects, this reference does not give an account of its full meaning—which is also constituted by something else, something that depends on the subject using the sign.

A question remains regarding Schopenhauer’s distinction of content and extent: which of these two, if any, should be understood as the “meaning” of a concept? It has already been remarked that Schopenhauer avoids using the term “meaning” when referring to concepts. In the above quotation he does not say that abstract concepts are meaningless but instead uses the figurative expression “emptiest and poorest.” However, some passages can be found where Schopenhauer seems to imply that it is the content (“Inhalt”) of a concept that constitutes its meaning: “concepts obtain all meaning (‘Bedeutung’), all content (‘Inhalt’), only from their reference to representations of perception, from which they have been abstracted, drawn off, in other words, formed by the dropping of everything inessential” [11, p. 474]. Here he seems to treat both terms as synonyms. Still, the possibility should be noted that in this passage Schopenhauer does not rigidly use the terminological differentiation of content (“Inhalt”) and extent (“Umfang”) he established in other passages.

Notwithstanding the uncertainty as to whether the content (“Inhalt”) of a concept can be unhesitatingly identified with its meaning, it has to be underlined that, for Schopenhauer, both the extent (“Umfang”) and the content (“Inhalt”) of concepts are rooted in perception. Thus, his distinction is founded upon a representational theory—both the qualities a concept connotes as well as the objects it denotes are representations of something that can be found in empirical reality. Otherwise the concept is empty.

In addition, it has to be noted that when Schopenhauer mentions concepts and refers to spherical diagrams indicating their extension it is difficult to make out whether he means signs of language or notions. It appears that this is irrelevant for him since he claims that it is impossible to think of a specific notion without having a sign for it (cf. [13, p. 244], quoted in Sect. 3). What can be theoretically distinguished—sign and notion signified by it—is almost incomprehensible from the point of view of a phenomenological analysis of the content of one’s mental processes—a perspective Schopenhauer assumes when analyzing conceptual thinking.

6 Use Theory of Meaning as Unnatural Way of Language Acquisition

Let us sum up the findings of this article. In Sect. 2 I showed that the classification of Schopenhauer’s theory of language and meaning depends on whether we take words or concepts as the carriers of meaning. In Sect. 3 I showed that the distinction between words and concepts is not strict in Schopenhauer’s philosophy and that under “concepts” he sometimes understands notions which cannot be considered elements of language. In Sect. 4 I showed that this ambivalence results in traces of both the representational theory of language and use theory of meaning appearing in Schopenhauer’s theory of language. In Sect. 5 I also showed that within the scope of the representational theory of language Schopenhauer seems to differentiate between sense and reference. Finally, in this section I would like to show that even though the representational and use theories seem to co-occur in Schopenhauer’s reflections on language, he himself clearly gives precedence to the representational theory.

As has been pointed out in Sect. 5, by the term “meaning” Schopenhauer sometimes means notion-like extralingual mental states of the subject that are represented by signs of language. Let us once again go back to Lemanski’s article in which he points out that Schopenhauer anticipates the use theory of meaning in a quotation from his Lectures (cf. [5, pp. 185–190]). In this quotation Schopenhauer clearly states that the “true meaning” (“wahre Bedeutung”) of a word, i.e., the “concept the word signifies” (“Begriff […], den das Wort bezeichnet”) can be abstracted from the “different context in which the word is found” (“aus dem verschiednen Zusammenhang[,] in dem man das Wort findet”). This is preceded by the statement that the “true value of words of a foreign language” (“den wahren Werth der Wörter einer fremden Sprache”) is acquired “ex usu” [5, p. 186] (cf. [13, p. 246]).

From these statements Lemanski infers that Schopenhauer hereby formulates the main thought of the use theory of meaning, which later will be formulated by late Wittgenstein as follows: “the meaning of a word is its use in the language” [5, p. 190] (cf. [17, para. 43]). Eventually, analysis of the quotation from the Lectures and other statements of Schopenhauer regarding meaning (that we have already discussed above) leads him to the thesis that the use theory of meaning and the representational theory of language need not be understood as opposing semantic theories given the fact that they appear simultaneously in Schopenhauer’s analysis of language and meaning [5, p. 190].

I hope to have shown in the preceding sections that the co-occurrence of these two semantic theories in Schopenhauer’s writings is strongly stimulated by his terminology, which does not strictly separate signs of language from their meanings. Consequently, this leads to his formulation of two different theories of meaning: (T1) the meanings of words are mental states and (T2) the meanings of concepts are the real objects they represent. The first theory enables the formulation of the use theory of meaning, just as Lemanski has shown, by providing the assumption that meanings are actually some kind of subjective mental states that can be inferred by the subject from observing how other people use certain words. Conversely, theory (T2) is a representational theory of language since in it meaning is understood as objects represented by concepts.

Now I would like to enhance this by showing that Schopenhauer brings the two theories into a hierarchy and clearly states which one he considers more important, again with reference to the quotation from the Lectures discussed by Lemanski. In this passage Schopenhauer clearly speaks about the acquisition of a foreign language, i.e., he refers to a situation in which the subject has already developed its concepts and has generally acquired the ability to use them. This situation is quite different from that of a child learning his or her first language. The child does not know any language and starts conceptualizing the world and referring to it with language for the first time, i.e., he or she develops what is to become the meaning of the words by using his or her native language. Interestingly enough, Schopenhauer actually elaborates on this issue in his late work Parerga und Paralipomena, in para. 372 at the beginning of Chapter 28 entitled On education, in which he distinguishes between two ways of acquiring concepts by a child. He calls the first one “natural education” (“natürliche Erziehung”) and describes it as the process by which a child comes in contact with empirical reality and as a consequence develops concepts that refer to facts from reality. He calls the second one “artificial education” (“künstliche Erziehung”) and describes it as the process by which a child is taught a bundle of concepts (and by “concepts” he means signs of language) that are not and cannot be immediately brought into direct reference with the empirical world. This can occur only after the signs of language have already been acquired. It is presented as a lengthy and tiresome process which, until finished, enables a lot of misinterpretations, mistakes, and misunderstandings⁸ [10, pp. 562–563].

It seems that in describing these two modes of language acquisition Schopenhauer actually refers to two semantic theories. Natural education is founded on the representational theory of meaning, whereas artificial education consists in acquiring concepts by means of the use theory. What Schopenhauer does here is in fact making a descriptive statement that

• (D) concepts are acquired both by means of empirical interaction with the world and by the interaction with the language others use,

and making a normative statement that

• (N) concepts should be acquired by means of empirical interaction with the world.

Obviously, (N) implies that the representational theory should be applied, whereas (D) presents the use theory as the one that is used, but should not.

It seems that Schopenhauer is exploring artificial education according to (D). Given the additional fact that the subject learning a foreign language already has its own concepts, it needs to infer the concepts behind the foreign words (i.e., their meanings) from how they are used in the foreign language. During this process the subject also realizes that in the foreign language there are signs that signify notions different from those for which there are words in its native language. Thus, it is neither sufficient nor even possible to simply learn the equivalent terms from the foreign language by using a dictionary since the signs of different languages signify different notions (cf. [13, p. 246]).

However, keeping in mind the interpretation of concepts as extralingual notions, we might also conclude that Schopenhauer is referring to two completely different processes: gaining the capacity to use any language and learning a second language. The first process, which we might call the conceptualization of the world, is presented with reference to the representational theory, whereas the second one refers to the use theory.

7 Conclusions and Prospects

Schopenhauer’s theory of language can be summed up with this sentence: signs of language represent private notions of subjects, which in turn represent real objects. This grounds the interpretation that it anticipates both the representational and use theory of language. Perhaps the most interesting thing to learn here is that these theories do not have to be understood as opposing. In reference to Schopenhauer’s influence on Wittgenstein, this provides an additional argument in favor of Bryan Magee’s stance that Schopenhauer’s philosophy might be interpreted as the common framework for early and late Wittgenstein, which would support the theory of continuity in his philosophy (cf. [6, pp. 324f.]). At least we can say that both theories of language represented by Wittgenstein can be traced to Schopenhauer’s observations on language—regardless of the fact that they might stem from Schopenhauer’s ambivalent use of terminology.

Based on Schopenhauer’s elaborations on education, we might conclude that language in the broader sense is a mode of representing the empirical world, but we acquire many of its elements by observing how other people use it; we infer the meanings of words from context and finally might refer them to reality. This way of acquiring signs of language is especially present when we are learning a foreign language. It seems that Schopenhauer realized the difference between the process of acquiring the ability to use language (i.e., establishing the connection between language and world and being able to express one’s thoughts in language) and the process of learning a foreign language—where the ability to use language is already apparent, but new signs of language and constructions have to be acquired.

Let us now synthetically sum up the article’s main findings.

• 1. Schopenhauer uses the terms “word” and “concept” ambivalently. In his investigation of meaning the anticipation of two different semantic theories can be traced: the representational theory of meaning and the use theory of language. The main reason for this is that he defines the meaning of words and concepts differently, and, at the same time, he occasionally uses the terms “word” and “concept” as synonyms.

• 2. These two semantic theories, although they are both present in his system, are not treated equally. He gives obvious and strong privilege to the representational theory of meaning. This can be seen in his claim that the entire content of concepts is derived from empirical experience as well as his normative claim that natural language acquisition should be rooted in empirical experience. The use theory of language is only discussed by Schopenhauer when he enumerates situations of unnatural language acquisition, such as when a child is forced to learn concepts it does not understand and when the phenomenon of foreign language acquisition is explained. Therefore, we might treat the use theory in Schopenhauer’s writings as an auxiliary theory.

• 3. In certain contexts, Schopenhauer seems to use the term “concept” in the sense of something that cannot be considered a sign of language at all (an unconscious notion).

Finally, there are a number of questions that could not be investigated in this article and need further research.

• 1. Schopenhauer assumes that every person’s use of language is based on a different conceptualization, since everyone gains their concepts from dealing with their own personal empirical experience, which, in fact, is private (cf. footnote 6, Sect. 2). This should consequently lead him to claim that the meaning/content of concepts is private. Instead, he seems to assume that each language community has a different conceptualization of the empirical world which can be realized when we learn a foreign language and see that the concepts behind its words are not equal with the concepts we hitherto used when speaking our native language. This means that he assumes that concepts have an intersubjective content within one language community. Thus, the question of whether and how he explains the possibility of this should be investigated.

• 2. In Sect. 5 it was shown that when discussing the content of concepts Schopenhauer distinguishes two dimensions of how concepts refer to reality, the content (“Inhalt”) and the extent (“Umfang”). Also the claim has been made that these resemble the distinction of sense and reference by Frege. This claim needs further investigation regarding both the systematic similarities between these philosophical terms, as well as the possible influence of Schopenhauer’s theory of language on Frege and his philosophical successors.