Hermann Cohen’s little book of 1871 took neo-Kantianism in a new direction, moving it away from a psychological and towards a more logical conception of epistemology. But Cohen was not alone in pushing neo-Kantianism in this direction. In the early 1880s another thinker appeared on the scene to give this movement even more impulse and energy: Wilhelm Windelband (1848–1915).
Although best remembered today for his work on the history of philosophy,1 Windelband was a major thinker in the history of neo-Kantianism. His significance for this movement is just as great as Cohen’s. Just as Cohen became the leader of the Marburg school, so Windelband became the father of the Southwestern or Baden school of neo-Kantianism. Windelband was the teacher of Heinrich Rickert (1863–1936), who was in turn the teacher of Emil Lask (1875–1915). Together, Windelband, Rickert and Lask form the inner core of the Southwestern school.
In the 1870s and 1880s, Cohen and Windelband seemed to be working together in shaping a new conception of epistemology, and in forging a new understanding of Kant. Both were sharp critics of the Friesian tradition and Helmholtzian programme, which would base epistemology on psychology. Both were champions of a new conception of transcendental philosophy, according to which its chief focus should be the quid juris? rather than quid facti?, the reasons rather than causes of knowledge. Hand-in-hand with this new understanding of epistemology went their new interpretation of Kant. Cohen and Windelband understood Kant’s philosophy as “critical idealism”, that is, as a strictly immanent analysis of the conditions of experience which involves no transcendent dimension, whether that of the thing-in-itself or Platonic forms. Because of their leading roles in the neo-Kantian movement, Windelband’s and Cohen’s views on these issues became the new orthodoxy of the 1880s and 1890s.
These affinities between Windelband and Cohen raise the question of their personal relationship. Although they were virtual contemporaries, the two men never seem to have met, though they had once clashed over academic politics.2 Because of the destruction of his correspondence, Cohen’s opinion of Windelband will probably remain forever unknown. We do know, however, Windelband’s opinion of Cohen, because in a late article discussing recent developments in contemporary philosophy the elderly Windelband found the fitting occasion to pass judgement on his slightly older counterpart.3 Windelband appreciated Cohen’s role in defeating the psychologistic tradition and in stressing the more logical side of Kant’s philosophy. It was Cohen’s “great contribution to have again validated the strict rationality of the Kantian philosophy”. But Windelband thought little of Cohen’s commentaries on Kant, which had made “the obscure more obscure, the difficult more difficult, and the complicated more complicated.” He also thought that Cohen had failed to get beyond the narrow Kantian paradigm of science, which limited it to mathematics and natural science; never did Cohen attempt to explain how there could be a science of history or culture. Whether Cohen’s emphasis on the method of the infinitesimal as the model for science would prove fruitful in the natural sciences Windelband did not venture to say; but he was confident that this would prove to be a “straitjacket” for philosophy.
Windelband was a very different writer from Cohen. While Cohen wrote a systematic philosophy, modelled around Kant’s three Kritiken,4 Windelband never came close to a system of his own. Although he believed in the value of systematic philosophy, he did not find the occasion, opportunity or energy to write one. His chief philosophical legacy is fragmentary. It consists in his collection of lectures and essays, which were first published in 1884 under the title Präludien.5 The final edition consists in essays and lectures spanning some forty years, from the late 1870s until the end of the first decade of the 20th century. The title was apt: it was Windelband’s way of admitting that his work was more suggestive than complete, more tentative than final. For these preludes, there would be no symphony.
With the benefit of hindsight, it is safe to say that Windelband made two central contributions to the neo-Kantian movement: his investigation into the logic of history; and his conception of philosophy as a normative enterprise. The concern with history will stimulate the later investigations of Rickert and Lask;6 and the normative conception of philosophy will be the basis for the philosophy of value, which will be a central interest of German philosophy in the later 19th and early 20th centuries.7 Windelband’s normative conception of philosophy remains one of the vital links between neo-Kantianism and contemporary philosophy, which has been no less concerned with the concept of normativity.8
The sources and origins of Windelband’s neo-Kantianism are obscure. There are no published letters or documents that show when, where, how, why and because of whom, he became converted to the critical philosophy.9 Kuno Fischer seems to have been the decisive influence. Windelband had been a student at Jena, where he heard Fischer’s lectures on Kant. The influence of Fischer upon him can be inferred from the importance he gave to Fischer, whom he regarded as the decisive figure in the birth of the neo-Kantian movement.10 Otto Liebmann also appears to have been important for Windelband’s neo-Kantianism. In a late tribute he praised Liebmann for being the first scholar to develop a genuinely critical interpretation of Kant, one neither psychological nor metaphysical.11 Windelband met Liebmann in Straßburg in 1882, after which they were in frequent correspondence.
Having treated Windelband’s work on the logic of history elsewhere,12 the focus of this chapter will be upon Windelband’s other contribution to neo-Kantianism, his normative conception of philosophy. We will see in the first five sections the precise meaning and problems of that conception, and then in the final three sections the stages in which it took shape in Windelband’s thinking. Windelband’s intellectual development was no less twisted and turbulent than Cohen’s. It was only after a long inner struggle that he arrived at his normative conception of philosophy. We will not consider, however, Windelband’s intellectual development after he formed the outlines of his normative conception. That would be a much longer story, which we have no space to tell here.
Windelband first put forward his normative conception of philosophy in a lecture he gave in 1881 on the occasion of the centenary of the publication of the Kritik der reinen Vernunft.13 This lecture, entitled simply ‘Immanuel Kant’, reveals the rationale for his neo-Kantianism as well as his normative conception of philosophy. To celebrate the centenary with a fitting tribute, Windelband proudly declares Kant’s Kritik der reinen Vernunft to be “the Bible”, or more literally “the founding book” (Grundbuch), of German philosophy (114). Much of the rest of his lecture is an attempt to explain why this is so, in both a historical and philosophical sense.
As the occasion required, Windelband ascribes great historical significance to Kant’s philosophy. There are, we are told, only two basic philosophical systems: that of Kant and that of the ancient Greeks (viz. Plato and Aristotle) (117). With Kant’s philosophy, there began “a whole new realm of thought”, one which stands in marked contrast to that of the Greek world. Kant’s philosophy was indeed a revolution, because, for the first time in Western culture, it broke with the legacy of classical Greece. Kant overthrew the intellectualism of Greek culture, which rested upon its naive faith in reason, its bold confidence that thinking alone could reveal the very essence of the world. Kant questioned this faith in reason in two ways: by teaching that reason cannot get beyond the limits of experience; and by separating morality and aesthetics from reason (in its theoretical sense) (121–122). Prima facie this seems to be a false account of philosophical history, because it seems to ride roughshod over the ancient Greek sceptics, who questioned faith in reason no less than Kant. But Windelband’s claim on behalf of Kant’s originality is not that he was the first sceptic, but that Kant and the Greeks represent “the two great philosophical systems”. The sceptics, being “intellectual nomads”, as Kant called them, have no resting place or system. Kant’s system of philosophy is not sceptical, Windelband stresses, because its aim is to determine the basic principles of all science, art and morality, the foundations on which all those disciplines rest (122). Kant’s system differs from the Greeks because it places these foundations in a different place than they: not in contemplation of being as being, still less in the intuition of the forms, but in experience and the autonomy of reason.
Part of the originality and importance of Kant’s philosophy, Windelband adds, lies in its close alliance with the modern mathematical sciences. The first important intellectual development since the Greeks, he tells us, was the rise of these new sciences (116). The purpose of Kant’s critique was to determine their foundations, to show the conditions for the possibility of a mathematical knowledge of nature. Hence Kant, entirely in the spirit of the new sciences, had adopted mathematization as the precondition of science itself: “I maintain that in each doctrine of nature there is only so much science to be found in it as there is mathematics.”14 Of course, the Greeks too were firm believers in the mathematization of nature, and they were indeed the founders of this method. But Kant differed from them in one important respect: he limited mathematical knowledge of nature to appearances. Mathematics was a human construction, the creation of our forms of sensibility, and as such not the reason according to which God created the world.
Although Windelband stresses the close connection between Kant and the new sciences, he is also quick to distance Kant’s philosophy from the new positivism (123). The positivists too wanted a close connection between philosophy and the sciences–so much so, that they wanted philosophy to disappear into them. That positivist elimination programme presupposes, however, their very naive conception of science. The positivists are naive about science, Windelband argues, because they do not recognize the intellectual preconditions for its possibility (123). They are simple empiricists who see knowledge as a collection of facts, as if these facts were simply given, and as if human sensing and thinking played no role in their constitution. It is one of Kant’s great contributions to have shown that there are intellectual preconditions of science. We cannot determine any point in time, undertake any measurement, calculate any weight, or conduct a single experiment, without applying basic concepts and presupposing fundamental principles. The very presence of these concepts and principles means that philosophy cannot be reduced down to the sciences, because it takes a philosopher to bring them to self-consciousness and to investigate their possibility.
The core of Kant’s philosophy, Windelband tells us, centres around the question how a priori representations relate to their object (134). That, at any rate, was Kant’s first formulation of his problem in his famous 1772 letter to Marcus Herz.15 But it is important to note, Windelband advises us, that Kant’s later thinking about this question changed its very meaning. As first posed, the question seems to assume that: 1) there is some object existing independent of our representations; and that 2) knowledge consists in the correspondence between them and their object. But it was these very assumptions that Kant overturned in the course of his investigation. Kant explains the possibility of knowledge not through the correspondence of a representation with an object but through the conformity of representations with rules (135). He replaced the concept of an object with that of a rule. For Kant, then, objects turn out to be nothing more than “determinate rules for uniting representations” (137).
It is in this context that Windelband introduces his concept of a norm (138). For him “a rule” and “a norm” are essentially synonomous terms.16 The analysis of truth into rules was one and the same as its analysis into norms. The new normative conception of truth appears explicitly in the general principle: “Truth is normativity of thinking” (138). Windelband did not invent the concepts “norm” or “normativity”, for which there was ample precedent in the writings of Lotze, Herbart and Sigwart. What is original to, and characteristic of, Windelband, however, is the central importance he gives to the concept of normativity in philosophy. Philosophy was for him essentially a science of norms. The task of philosophy is to determine the basic norms that bestow value upon all human activity, whether thinking, willing or feeling (139). There are basic norms not only in science but also in morality and art. There are then three parts of philosophy: science, ethics and aesthetics.
What, though, are these norms? What ontological status do they have? Where do they exist? Where do they come from? Such questions, Windelband replies, Kant wisely refused to answer. “Kant declined any metaphysical interpretation of these rules” (137). The attempt to answer such questions was the enterprise of speculative idealism. But Windelband issues an embargo against all such speculation, insisting upon the strictly immanent status of Kant’s philosophy, whose intent is to stay firmly within the limits of experience.
In forgoing all metaphysical speculation, Windelband was self-consciously and deliberately pushing philosophy away from its traditional function as a worldview. He tells us explicitly and emphatically that Kant’s task was to provide not “a worldview” (Weltbild) but a theory of “normal consciousness” (das normale Bewußtsein), that is, consciousness insofar as it conforms to norms (141). Yet on no other issue did Windelband vascillate so much as the need for and value of a worldview. As a student of Lotze, he had grown to believe in the necessity of metaphysics, which, in his early years, he makes an integral part of his epistemology.17 In his later years, however, he will come full circle, reaffirming the need for and value of metaphysics as one of the legitimate concerns of philosophy.18 What indeed distinguished Kant’s philosophy from a soulless positivism, Windelband would later teach, is precisely that it provides a complete worldview. With the later Windelband, “Kant’s worldview” would enter into the discourse of the Southwestern school.19
The ultimate upshot of Kant’s philosophical revolution, Windelband reckons, is that it overturned the traditional way of conceiving knowledge as “the mirror of nature” (Spiegel der Welt) (127). According to that ubiquitous metaphor, representations are images of objects which they resemble, just like a mirror image resembles its object (123). We determine whether representations are true or false by comparing them with their prototypes, which are objects existing independently from them. The Kantian Revolution means, however, that the standard of knowledge is not in the object but inside the subject itself, residing in the basic norm for organizing its representations.
Such, in a nutshell, was Windelband’s first statement of his normative conception of philosophy. For a 21st-century reader, it is remarkable to find Windelband explaining Kant’s philosophical revolution in terms of replacing the metaphor of the mirror of nature with that of normativity. Since Richard Rorty, anglophone readers have grown accustomed to seeing that transformation taking place in the early 20th century with philosophers like Dewey, Wittgenstein and Heidegger.20 But if Windelband shows us anything, it is that we have to place that transformation much earlier: with Kant, or at least with his interpretation of Kant.
The conception of philosophy Windelband formulated in ‘Immanuel Kant’ was his solution to the obsolescence crisis which had troubled all neo-Kantians. Just how it resolves that crisis Windelband did not explain in his lecture, but he turned to that task only a year later in a long essay, his 1882 ‘Was ist Philosophie?’21
Windelband’s account of the origin of this crisis differs in important ways from his predecessors. For Windelband, it was not the result of the collapse of speculative idealism in the 19th century, as it was for Zeller and Meyer. Rather, it began with the general scepticism about metaphysics in the 18th century. After philosophy declared its independence from theology during the Renaissance, it conceived its characteristic task to be metaphysics, a universal theory of being or “the general science of the cosmos” (die Gesamtwissenschaft vom Weltall). Each of the new empirical sciences lay claim to knowledge of a special part of the world; but philosophy found its unique calling in metaphysics, which would be a knowledge of the world as a whole (16). Philosophers of the 18th century, however, became very concerned with epistemological issues, with questions about the origins, powers and limits of knowledge; and the more they examined the aims and claims of metaphysics, the more they found them wanting (18). The crisis of philosophy then arose when metaphysics could no longer legitimate itself, when it could not prove its worth as a science. But if metaphysics is bankrupt, what is left for philosophy? There seemed to be nothing more for it to do. The entire realm of nature had been carved up by the particular sciences. Windelband summarizes the predicament of philosophy like this: “Philosophy is like King Lear, who gave all his goods to his children, and who now must be thrown on the street like a beggar.” (19)
Though it seems purely historical, Windelband’s account of the obsolescence crisis makes an important point against his neo-Kantian predecessors. Unlike Fischer, Zeller and Meyer, Windelband does not think that the solution to this crisis lies simply with epistemology; for him, epistemology is part of the problem rather than the solution, because it created the crisis by undermining metaphysics. Windelband sees epistemology, reflection on the powers and limits of knowledge, as a pursuit characteristic of pre-Kantian epistemology; and he is perfectly clear that philosophy cannot resolve its crisis if it becomes “a theory of knowledge”, that is, “investigation into the nature of science, the process of knowledge” (18). This means that the neo-Kantian proposal to make philosophy into epistemology cannot be sufficient; we have to go on to specify something about the kind of epistemology. Though Fischer, Zeller and Meyer had noted Kant’s debts to the epistemological tradition of Descartes, Locke and Hume, they saw his differences from them simply as a matter of greater focus and emphasis. But Windelband does not think that this goes far enough; it fails to sees a fundamental difference in kind between pre-Kantian and Kantian epistemology. To miss this difference has profound consequences: ignorance of the very nature of philosophy itself.
For Windelband, epistemology before Kant was essentially a psychological or historical enterprise. Its fundamental task was to explain the causes of knowledge (22). It would explain knowledge as it would any fact in the world: by deriving it from its causes or general laws. Here Windelband seems to have in mind especially Locke’s “plain, historical method” or Hume’s “science of human nature”. But Kant broke profoundly with this tradition, Windelband explains, because he saw epistemology not as a psychological or historical investigation into the causes of knowledge but as a critical investigation into its value or validity (23). What was at stake was not the causes of representations but the foundation or justification of judgements (25). There is a fundamental difference in kind between these concerns, Windelband argues, because both true and false judgements have natural origins; both arise from a process of natural necessity and fall under natural laws. These natural laws do not provide, therefore, a criterion for the truth of judgements; so a complete explanation of their causes will not answer the question whether the judgement is true (23). It is solely this question, however, that is the special province of epistemology: “for the theory of knowledge it is only a matter of whether the representations are valid, i.e., whether they should be recognized as true.” (25)
Implicit in Windelband’s account of pre-Kantian epistemology there lay another critique of his neo-Kantian predecessors. Windelband was in effect saying: Fries, Herbart, Beneke, Helmholtz, Fischer, Zeller, Meyer and Lange have all misunderstood the essential nature of the Kantian enterprise. They have misunderstood it because, like their pre-Kantian antecedents, they have misconstrued it in psychological terms. They have failed to notice the basic distinction between explaining the causes of beliefs and assessing the reasons for them. What was fundamental about the Kantian Revolution in philosophy, Windelband maintains, is precisely this shift away from causes and explanation towards reasons and criticism. Although this criticism of Windelband’s neo-Kantian predecessors is completely implicit in “Was ist Philosophie?”, it became entirely explicit in a later work, Die Philosophie im deutschen Geistesleben des XIX. Jahrhunderts, where Windelband takes a long retrospective look at all the philosophical movements of the 19th century, not least the neo-Kantian movement, which he expressly takes to task for its psychological interpretation of Kant.22
The central argument of Windelband’s “Was ist Philosophie?” is that only his normative conception of philosophy can overcome the obsolescence crisis. If we define epistemology in terms of psychology in the manner of my noble neo-Kantian predecessors, Windelband was saying, then we do not escape the crisis but remain caught in its snares. For, by the late 19th century, psychology had become a natural science of its own, independent of philosophy.23 It now applies the same methods of observation and experiment as physics and physiology, and it now employs the same paradigms of explanation as these sciences by subsuming psychic events under general laws. Whatever Herbart might say, psychology gets verifiable results independent of metaphysics, which is not needed to guide its enquiries. But if psychology is now an empirical science like any other, then those who fail to distinguish epistemology from psychology virtually ensure the obsolescence of philosophy. Rather than being a distinct discipline apart from the empirical sciences, philosophy will just become one among them. Fries, Herbart and Beneke had all dreamed of psychology becoming a science. But now that their dream had become a reality, where does that leave philosophy? Their psychologistic answers to that question only leads to the obscolescence they are so eager to avoid.
Despite these criticisms of the neo-Kantians, Windelband still accepted the common neo-Kantian conception of philosophy as epistemology (Erkenntnistheorie); its just that he understands epistemology in very different terms. It is not the explanation of knowledge according to the laws of psychology but the criticism of knowledge according to the standards of reason. More specifically, philosophy is the criticism of particular kinds of claims to knowledge, namely, synthetic a priori or universal and necessary claims. Philosophy is therefore “the critical science of universally valid values” (29). There are three domains where universally valid claims are made: in science, morality and art, corresponding to the realms of cognition, volition and feeling; and so there is accordingly a philosophy of science, morality and art. In a marked departure from his own earlier views, whose causes we will examine later, Windelband warns against confusing epistemology with two very kindred but still very different disciplines: psychology and metaphysics (28). The problem with conflating philosophy with these disciplines is that one reifies or hypostasizes values, making them into psychological processes or archetypes of reason. Psychology and metaphysics are theories about what is or must be; but as a science of value philosophy concerns what ought to be.
To define further the nature of philosophy, Windelband attempts to specify its precise subject matter. He makes a distinction between two different kinds of propositions: judgements (Urtheile) and appraisals (Beurtheilungen) (29). Although they have the same grammatical form, where a predicate is attributed to a subject, these propositions have a very different meaning. In a judgement we attribute a property to an object; but in an appraisal we add nothing more to the representation of an object (30). Appraisals concern not the object itself so much as the subject’s attitude towards it; they are essentially acts of approval or disapproval (30). The task of an appraisal is therefore to determine the value of the object. This distinction is crucial for defining philosophy, Windelband maintains, because its special concern is appraisals (32–33). Philosophy does not describe appraisals, like the historical sciences; and it does not explain them, like psychology; it does not treat them as if they were objects in the world at all. Rather, it actually makes appraisals, determining what should be or have a value.
When a critical philosopher makes an appraisal, Windelband explains, he assumes that something should be recognized as valid for everyone; he makes, in other words, a universal and necessary claim to value. In doing so, he presupposes “a universal criterion that should hold for everyone alike” (37). It is the task of the critical philosopher to determine these criteria, to ascertain the highest standards of judgement in the fields of science, morality and art. These criteria are norms, laws or ideals about what ought to be; they have no reality themselves because by them we measure or appraise all reality. In each field we presuppose the concept of an ideal being who recognizes and follows these norms. This ideal being Windelband calls “normal consciousness” (Normalbewußtsein) (45). We do not assume that this normal being exists, only that it ought to do so (44). Understood in these terms, philosophy is “the science of normal consciousness” (46).
Despite all Windelband’s efforts to clarify the task of philosophy, there remained major ambiguities about his programme. It is unclear whether the Windelbandian Wertwissenschaftler is making values, assessing things according to values already made, or simply describing values and pointing out the means to and consequences of them. Sometimes it seems that he is doing some combination of all three. Only the first is really an appraisal in the strict sense, because the other two tasks need not involve accepting the values on which the appraisal is made; one is really judging rather than appraising because one is simply drawing out the logical conclusion of holding a value, or one is pointing out certain empirical facts (viz. the means to ends, or the consequences of acting on a value). Windelband’s confusions on this score proved a troublesome legacy for Rickert and Weber, who struggled to fight through them.24
Windelband’s solution to the obsolescence crisis of philosophy in “Was ist Philosophie?” rested entirely upon his distinction between two very different intellectual domains, the normative and the natural. While the natural sciences concern themselves with the natural realm, determining laws about what is the case, philosophy deals with the normative realm, determining standards or ideals for judgement about what ought to be the case. The distinction has the advantage of clarity and firmness. It also seems to achieve precisely what it sets out to do: it makes philosophy autonomous, giving it a unique and separate domain apart from the empirical sciences. So far, so good.
Yet the very strength of the distinction is also its weakness: it makes such a strong and sharp separation that the two domains seem to have nothing to do with one another. If philosophy deals with norms about what ought to be the case, and if it is not at all concerned with laws about what is the case, then it seems doomed to irrelevance. Philosophers can talk and theorize all they want about their ideal world of truth, beauty and goodness, but if their ideals are not followed, if their norms are not realized in the actual world, their safe and secure domain turns into a fantasy world. Windelband, it seemed, had solved the problem of obsolescence only at the price of irrelevance.
No one was more acutely conscious of this problem than Windelband himself. Philosophy was never for him an entirely theoretical activity having no bearing on the world itself. It was an essential part of Kant’s revolution in philosophy, in his view, that it rejected the classical Greek conception of philosophy as contemplation, and that it would attempt to determine the fundamental norms behind all human activities. Philosophy should actually make a difference to the world, Kant held, because self-consciousness of these norms determines how we follow them, how we really think, act and feel in the world. Windelband shared this Kantian belief in the powers of philosophy; but the problem was how to justify it, how to explain the relevance of philosophy to life and the world. This was an especially pressing problem for Windelband, given his own sharp and fast distinction between the normative and natural.
Windelband’s first sustained effort to solve this problem is his 1882 essay ‘Normen und Naturgesetze’,25 which first appeared in the 1884 edition of Präludien. Though he would often return to this problem in later years, this essay would set the agenda for all his later efforts to deal with it.
Windelband first sets the problem in the context of one classical philosophical conundrum: freedom of will. This seems to make the original problem all the more insolvable by tying it to one of the most intractable issues in all philosophy. But Windelband felt he had no choice, that the deeper issue was unavoidable. The very possibility of normativity hung in the balance. For if all events in the natural world are governed by laws, such that they must occur and cannot be otherwise, what is the value of having norms? What is the point in telling people what they ought to do? Norms have a point only if we have the choice to act or not act on them; but if everything is determined by natural laws, so that we act necessarily in one way or another, there seems to be no choice. Windelband then generalizes the problem, stressing that the same problem arises for other norms, for those of science and art as well as morality. In science we assume that we have the power to follow the norms of logic; and in art we suppose that we have the capacity to follow the norms of painting, composing, building, etc. If all thinking and feeling were determined by natural necessity, these norms would be just as pointless as those of morality. So Windelband poses his problem in the most general form: What is the relationship between norms and laws?
To understand their relationship, Windelband advises, we must first have a clear account of the difference between them. We must grasp the different activities involved in each, in applying laws and in making norms (65). The main difference is simple enough. A norm is not a principle of explanation; and a law is not a standard or criterion of judgement. This is because laws explain things, whereas norms assess or appraise them. With laws, we determine what must be the case by subsuming particular events under general relations of cause and effect; and with norms, we determine what ought to be the case. Through a norm we determine what something should be if it is to be true, good or beautiful (66–67).
Once we put the distinction this way, Windelband thinks, we can see that there is no conflict between norms and laws. They are just two different activities which treat their object from complementary perspectives (67). Although they are very different, they are also connected, because the norm prescribes what we ought to do in the world; since ‘ought’ implies ‘can’, it is valid only if we can follow it, only if we can act upon it (68). The problem of accounting for the relationship between norms and laws is then to explain the connection between them while still recognizing their differences. We must avoid, Windelband warns, two extreme views: completely separating norms and laws, as if they have nothing to do with one another; and conflating them, as if there is really no difference between them (68).
What is the middle ground between these extremes? Windelband attempts to plot it by stressing the contingent connection between norms and laws. All particular human actions, as events in nature, conform to laws, and they occur of necessity. But it is contingent whether these actions happen to follow norms. They might comply with norms; but they also might not comply with them. Although the mere fact that an action occurs of necessity does not exclude its compliance with norms, it also does not entail it. For even actions that violate norms still follow natural laws (69).
Although norms and laws are only contingently connected regarding whether a particular action complies with a norm, there is another respect in which they are necessarily connected. Norms are intended to be followed, to be acted upon in the world, and they assume that we have the power to act on them. They presuppose that there are many possible ways of acting in the world, that there are different forms of thinking, acting and feeling, and they select one from them as the right or correct way or form (72–73). Thus the laws of logic are a selection from the possible forms of association in consciousness; the laws of ethics a selection from the possible forms of motivation; and the laws of aesthetics a selection from the possible forms of feeling (73). From all the possibilities of thinking, willing and feeling, norms prescribe one as the best or most worthy.
Because norms are intended to be acted upon, and because they are conceived as one way of thinking, willing and feeling in the world, they have for Windelband a necessary connection with the natural processes that realize them. A norm is for him the realization of value through a natural law, which is the means for the actualization of its end. Windelband then states the following definition: “Norms are those forms of actualization through natural laws which should be approved for the end of universal validity.” (74)
The close connection between norms and laws appears even more clearly, Windelband thinks, once we recognize that norms are supposed to be reasons for our actions. They are, as he puts it, the “determining grounds” (Bestimmungsgründe) of human conduct (85). When we accept the validity of a norm, we adopt it, under the appropriate circumstances, as the reason or justification for an action; it serves as a constraint upon us, telling us the particular thing that we should do among the wealth of possibilities. When we resolve to act upon the norm, it then becomes part of the process of law-governed events in the world (87). In this way ethical and aesthetic norms become not only reasons that justify actions but also causes that produce them.
For all their apparent plausibility, Windelband’s attempts to connect norms and laws raise serious difficulties. While it is indeed the case that a norm is conceived as a manner of action, it also cannot be simply defined in such terms. For there is more to a norm than any set of actual actions. A norm determines what we ought to do, and just because “ought” is not reducible to “is”, so the norm cannot be reduced down to any particular matter of fact, in this case manner of action. But, leaving that difficulty aside, there is another no less serious one in assuming that norms are actualized through natural laws. For this begs the question whether we really can act on these norms, and whether they really can be actualized in a deterministic universe. If all actions occur of necessity, such that they cannot be done otherwise, then the norm becomes pointless. We are not acting on the norm but just following a natural law. Windelband has still not addressed, in other words, the problem of freedom on which normativity rests.
Windelband’s solution to this second difficulty is his compatibilism regarding the grand question of the relationship between freedom and necessity. If an event happens according to natural laws, he argues, that does not exclude the possibility that the agent has adopted a norm as a reason for his action. We assume that natural necessity is incompatible with normativity only because we think that acting on norms involves the possibility of doing otherwise, where that possibility is defined in terms of acting countercausally, contrary to the laws of nature. But no such assumption is necessary, Windelband contends, because reasons can still be causes of actions even though they are determined according to natural laws. We need not parse phrases about doing otherwise in terms of some mystical power of acting countercausally because all that these phrases mean is that we would have done differently if we, or the circumstances, were different (88).
Given his compatibilist position, it should come as no surprise that Windelband is highly critical of Kant’s own solution to the problem of freedom (62, 88, 97). Kant’s thinking about this issue began from the false premise that moral responsibility and natural causality are incompatible, that we have a power of transcendental freedom that involves spontaneity, which is the power to act without any prior causes. In assuming that the will consists in such a power, Kant pushed the will into a spooky noumenal world, Windelband complains, where it is inaccessible to science as well as moral imputation. Hence, on Kant’s own admission, the real morality of our actions are unintelligible to us.26 Above all, we need to overcome Kant’s dualism, Windelband advises, because its radical separation between normativity and nature makes the connection between them obscure and mysterious. “The realm of freedom is in the middle of nature; it is that province where only norms are valid; it is our task and happiness to settle in this province.” (98)
Despite these criticisms of Kant, there was still one respect in which Windelband, like Liebmann,27 remained a very stubborn Kantian about the whole question of freedom. He was a Kantian of 1787 rather than one of 1792, in that he adopted Kant’s account of freedom in the second Kritik over that in his Die Religion innerhalb der Grenzen der bloßen Vernunft. Windelband accepted Kant’s account of freedom as autonomy, as the power to act morally. “Freedom is the domination of conscience. That alone that deserves this name is the determination of empirical consciousness according to the consciousness of norms . . . Freedom means to obey reason.” (87–88) Windelband’s attraction to this concept of freedom is readily understandable, because it connects normativity with laws and removes any power of doing otherwise that seems to contradict these laws. But there is a problem with this conception of freedom, a problem that Kant realized in the early 1790s and that Windelband seems to have forgotten: that defining freedom in terms of moral action makes it impossible for us to be free and immoral. To avoid this problem, Kant made his famous distinction between Wille and Willkür, between the will and choice, and stressed the importance of Willkür or choice for moral responsibility and freedom. But the price of such a doctrine is again making a dualism between the realms of freedom and nature.
Whatever their ultimate merits, it should be clear from the criticisms suggested here that Windelband was very far from a having an unproblematic account of the relationship between normativity and nature. This is hardly surprising, given that figuring out that relationship involved the classical problem of freedom. It was one of the merits of Windelband’s discussion of this relationship that he had never lost sight of that basic problem. But just because that problem was so complex and imponderable Windelband would return to it again in his later years. We will eventually have to consider his later attempt to address this conundrum of conundrums.
Windelband returned to the question of the vocation of philosophy in another article of the early 1880s, ‘Kritische und Genetische Methode?’, which was written in 1883.28 This was his most rigorous attempt yet to define the nature of philosophy and to provide it with a firm foundation independent of the empirical sciences. Windelband now approaches his problem in a completely new manner, attempting to distinguish philosophy from the empirical sciences through their different methods. Kant had already discovered the proper method of philosophy, Windelband tells us, though he did not develop it with sufficient clarity. As a result, there has been constant confusion about how its method differs from that of empirical science. To prevent that confusion, Windelband strives to explain the distinctive method of philosophy in the most simple and fundamental terms. This article too had a polemical point: it is directed against all those neo-Kantians who wanted philosophy to imitate the methods of the empirical sciences.
Windelband casts his discussion in strictly logical terms, beginning with a detailed explanation of the difference between the inductive and deductive methods. Both these methods, we are told, presuppose certain basic principles or axioms. Kant would call them “synthetic a priori principles”, but to avoid the controversy surrounding Kant’s doctrine of the synthetic a priori, Windelband prefers the simpler term “axiom”. He then defines philosophy as the science of axioms (107–108). Logic determines the axioms of the sciences, aesthetics the axioms of art, and ethics the axioms of morality. The main task of philosophy is to expound the complete system of these axioms.
The chief problem of philosophy, Windelband informs us, is “the validity of axioms (die Geltung der Axiome) (108). Somehow, philosopy has to assess or evaluate axioms. But how can it do this? Because they are the basis of all proof, axioms cannot be proven themselves. We cannot assess or evaluate them, it seems, because to do that already presupposes them. Any attempt to justify them would be circular, assuming them in order to prove them. Still, Windelband reassures us, these axioms have an immediate certainty, a self-evidence, that does not require proof or demonstration. The business of philosophy is not to prove (beweisen) but to show (aufweisen) this immediate certainty, to lay this evidence before our eyes (109).
Philosophy can show this immediate certainty in two ways. Either one reveals the “factual validity” (tatsächliche Geltung) of these axioms by showing how they are embedded in the practices of everyday life, by revealing how they are involved in all our ways of thinking, feeling and acting; or one reveals their “teleological necessity” by showing how accepting these axioms is necessary to achieve certain ends (109). The former procedure is the genetic method; the latter is the critical method. For the genetic method, the axioms are actual ways of understanding, feeling and acting; for the critical method, however, they are not facts at all; they are norms. Norms prescribe means to ends; they tell us what we should do so that our thinking is true, our willing is moral, and our feeling grasps the beautiful (109). The ultimate justification for them is that they prove to be efficient and necessary means of attaining these ends.
Although Windelband is not so explicit, it is clear that his distinction in methods presupposes his earlier distinction in ‘Normen und Naturgesetze’ between norms and natural laws. The central concern of the genetic method is to determine a specific kind of law, namely, those which explain why people adopt axioms on psychological and cultural-historical grounds. The main interest of the critical method, however, is to determine a specific kind of norm, namely, that which prescribes how we achieve certain ends. The difference between these methods also follows along the lines of Windelband’s earlier distinction between judgement and appraisal. All the ambiguities about appraisal—whether it is making or prescribing values, assessing things according to values or simply describing values—hold for Windelband’s critical method.
However we describe the critical method, Windelband thinks that it alone is characteristic of philosophy. The genetic method falls within the provinces of the empirical sciences. All those who want to push philosophy in the direction of the empirical sciences assume that it has a genetic method. But if we take philosophy in this direction, Windelband argues, we pay a terrible price: forfeiting the validity of axioms, their claim to universal and necessary worth (114). The genetic method shows only that axioms are “factually valid”, that is, it shows how people adopt them in the practices of their ordinary life, how they as a matter of fact assume them in the sciences, arts and morality. But it is an old and basic logical truth: what happens as a matter of fact cannot be the basis of a universal and necessary principle. What is believed or assumed is so only now and then by this or that person; and even if everyone believes them all the time, it still does not prove that they are true. The genetic method ends with a causal theory about the origins of beliefs, showing them to be the necessary result of psychology and cultural history; but this causal necessity proves nothing about validity for the simple reason that it applies to invalid as well as valid beliefs (115). For the genetic method, all beliefs are on a par, all are equally necessary results of psychological and historical causes. Thus those who push philosophy in the direction of the empirical sciences, Windelband concludes, have to accept a very problematic consequence: a complete relativism (116).
Relativism, of course, is the great challenge to Windelband’s conception of philosophy. If relativism is tenable, then there is no philosophy, no science of universal and necessary axioms, all for the simple reason that there really are no such axioms. Windelband, however, found it hard to take relativism seriously. He argues that a radical relativism is not sustainable because it is self-refuting, presupposing the very axioms that it attempts to bring into question (116). The relativist must assume, for example, that it is possible to ascertain certain basic facts (viz. cultural diversity), that it is possible to draw general conclusions from them (viz. that all principles are valid only for a specific time and place), and that these have a universal validity; but in assuming all this the relativist presupposes the very rules he disputes. As Windelband puts it: “The more the relativist tries to prove his doctrine, the more ridiculous he becomes; for he refutes all the more what he wants to prove.” (117)
All this was much too quick and dismissive, however. Self-refutation is indeed a danger for those who would dispute the basic rules of logic. But what about the moral and aesthetic relativist? Where is the self-refutation in denying any proposed norm of morality and aesthetics? To deny them might be a case of bad behaviour and bad taste, but hardly logical absurdity. It is indeed questionable whether Windelband’s critical method avoids relativism any more than the genetic method. The critical method attempts to show how certain axioms are means to ends; but what if one does not accept these ends themselves? What about the immoralist, who does not want to live according to universal laws? What about the dada artist, who does not want to feel beauty? What is to be said against these people? Here Windelband is at least clear and frank: there is no in talking with them (123). The presupposition of the critical method, he admits, is the faith that there are universally valid ends. Whoever does not share that faith should simply “stay at home”; the critical philosophy can do nothing for him (123).
The most serious difficulty facing the critical method, Windelband acknowledges, is how the individual comes to know the standpoint of normal consciousness. There is a danger, which Windelband fully admits, that the individual confuses his individual, empirical opinions, desires and tastes with universal and necessary norms (124). The same problem appears on a more general cultural level when a particular nation or people elevates its own mores into universal norms, as if they should be binding on all people. It was the very purpose of the genetic method to expose this kind of specious universality and necessity, to show how the apparently universal and necessary really arises from a particular people at a particular time. What is to prevent the philosopher from making these kinds of specious generalizations, from making his own private views, or the customs and fashions of his own culture, into universal laws for all mankind?
Despite these dangers, Windelband is confident that the critical philosopher, if he is only careful enough, can avoid this fallacy. Since his norms prescribe means to ends, his task is only to determine the necessary means for the realization of ends; and to do this, he does not have to refer to any data of his own individual or empirical consciousness (125). All that he has to show is that the means are really effective for achieving the ends. But this response too was very hasty. Granted, it is a straightforward matter to determine means to ends, it is so only once we know the ends; but this begs the question of how the critical philosopher is to know these ends in the first place. That was the very problem to be solved. Windelband seems to assume that nothing more is involved than knowing the general ends of science (acquiring knowledge), morality (doing the good) and aesthetics (feeling the beautiful). But this account is simply too abstract to be of much help, because the critical philosopher still needs to know in what, specifically, knowledge, goodness and beauty consists. It is precisely in attempting to provide a more specific account of knowledge, goodness and beauty that controversy arises.
An even greater problem arises in trying to solve this difficulty. Windelband admits that the critical philosopher cannot simply spin the universal and necessary principles of reason out of his own head. To know these principles, he has to see how they are embodied in particular ways of thinking, acting and feeling. Just as the linguist knows the rules of grammar from seeing how people actually speak, so the critical philosopher knows the norms of logic, aesthetics and ethics from seeing how people actually think, feel and act. As Windelband puts it: “Teleological construction requires not only the determination of ends, but also the consideration of the material in which these ends should be realized.” (127) But now we face the same problem that Windelband raised against the genetic method. Namely, from all these particular facts about actual thinking, feeling and acting we cannot draw universal and necessary conclusions. Windelband is very clear that “consideration of the material” should be only a means of discovering the principles of reason and not a means of justifying them (127). This is an important distinction to make, and we have seen how it had been applied with good effect by Fries and Meyer; but it does not help Windelband in this case. For it still does not tell us how from this material we can even discover the principles of reason. The problem is that there are many ways of thinking, feeling and acting; and which ways are the instances of the rules? The gap between “is” and “ought”, between natural laws and norms, once again lurks in the background.
In the course of thinking about how the critical philosopher knows the principles of reason Windelband began to wrestle with another thinker who had considered this very problem some eighty years before him: Hegel. Rather than condemning Hegel, as so many neo-Kantians before him, Windelband pays tribute to him. He praises Hegel’s “deep wisdom” in recognizing that rationality reveals itself only through history (133). There is no escaping the fundamental importance of history in knowing the principles of reason, Windelband realizes. The particular ways of thinking, willing and feeling that embody human rationality are intrinsically historical phenomena, appearing in a specific time, place and culture. The individual’s awareness of the principles of reason is also, as a psychological event, a phenomenon that takes place in history. Windelband makes an even greater concession to Hegel when he claims that history provides more suitable material to know the content of reason than psychology. While psychology informs us only about broad formal or structural features of the mind, namely, that it appears in willing, thinking and feeling, history shows us the actual content of reason, namely, what we will, think and feel (131–132). For all these reasons, Windelband concludes that “knowledge of all the content of rational values grows out of the critical illumination of history” (134).
Having paid his homage to Hegel and having recognized the importance of history for rationality, Windelband still fell short of adopting a genuine Hegelian position. While Hegel strived to overcome the gap between the normative and natural through his concept of reason in history, Windelband insists that there is, always was and forever will be, a gap between the normative and the natural, the rational and the historical. There is simply no catapulting over Lessing’s broad and ugly ditch, because it is impossible to infer universal and necessary validity from the facts of history. For Windelband, unlike Hegel, history is essentially contingent in the sense that none of its facts are logically necessary; history conforms to causal laws, to be sure, but not necessarily to logical ones, because the non-existence of its laws would not be self-contradictory. And so Windelband concludes: “For the critical method, the historical course of development, in its essentially empirical and contingent determinacy with respect to ‘the idea’, cannot have a systematic meaning.” (133)
Windelband’s interactions with Hegel would continue in his later years. The more he thought about “the old man”, the more he admired him. The attraction grew to such an extent that the older Windelband nearly became a Hegelian. It was Hegel who helped to convince him that history should be “the organon” of the critical philosophy (i.e. the instrument by which it comes to know the principles of reason),29 and that the critical philosophy should become a philosophy of culture.30 It was also Hegel who taught him that the philosophy of history is the means to overcome the dualism between the normative and natural, a goal that seemed more and more desirable to the ageing Windelband. His late philosophy of history is indeed pure Hegel, making the self-consciousness of freedom into the telos of history.31 In an article he wrote in 1910 on the nascent Hegel revival in Heidelberg, Windelband announced that, as long as they refrained from “historicism”, the neo-Hegelians had nothing to fear from the neo-Kantians.32 There was some sweet irony in this declaration of Kantian-Hegelian friendship, given how much the neo-Kantians once loved to hate Hegel. But at the dawn of the new century the old neo-Kantian animus had grown old and cold. A neo-Hegelian movement was in the making.33 The owl of Minerva had flown back to her home along the banks of the Neckar.
The fundamental business of philosophy, according to Windelband’s normative conception of this discipline, is to explain the possibility of norms. Philosophy has to be able to explain how it is possible for people to follow or act upon them. Nothing seems simpler, though, than the fact that people can act on norms, because they do this all the time. They follow instructions, they play games, they observe ancient customs. But, as happens so often in philosophy, what at first seems perfectly obvious later turns out upon reflection to be problematic, if not downright impossible. The possibility of following norms becomes problematic not least because they presuppose choice, that we might decide not to act on them, that we have the possibility of not following them; if we acted on them immediately, automatically and necessarily, it seems there would be no point in making them. Norms would be not imperatives but predictions. So, given that we must have the choice to follow them, the thorny question arises: How are norms possible in a deterministic universe?
We have already seen how, in his early 1882 essay ‘Normen und Naturgesetze’, Windelband raised and struggled with this question. He wanted to build a bridge between the normative and natural realms, and he made their connection dependent on the problem of freedom, which he had generalized so that it concerned all forms of norm following, whether the norms were moral, logical or aesthetic. But Windelband’s first attempt to deal with this question was at best sketchy and provisional, and so it became necessary for him to return to it in later years. His final and most systematic effort to deal with it is his Über Willensfreiheit, which consists in twelve lectures which he first gave in Heidelberg in the Winter of 1903.34 Because of the importance of freedom for Windelband’s normative conception of philosophy, we do well to examine the main ideas behind these lectures.
In his first lecture Windelband defines the nature of his problem and his strategy for tackling it. When we consider the question of freedom we should realize, he advises, that it is never just a matter of affirming or denying freedom of will (6). The concept has too many meanings and whether one affirms or denies it depends on the particular sense. We also need to keep in mind that freedom is a relative concept (11–12). We are never free absolutely but always in some respect. In applying the concept we need always to ask ‘free from what?’ We can be free in one respect and not in another. The question of freedom also depends on the particular kind under discussion. There are for Windelband three kinds of freedom, each of which corresponds to three phases of volition (17). The first phase is the genesis of desire or the act of will; the second is the choice between different desires; and the third is the action itself (18). The whole question of freedom divides accordingly into freedom of will, freedom of choice and freedom of action. Windelband’s discussion of freedom takes each of these aspects separately.
Windelband begins with freedom of action, the third and simplest form of freedom (19). Freedom of action is simply a movement of the body caused by the will, so that it is some purposive movement (19). We can separate, at least conceptually, will and action (20), Windelband maintains. This is because many of our reflex movements are involuntary, and many impulses of will do not result in action (20). Freedom of action, however, joins the will and action, so that it means simply that we do what we will (20–21). Freedom of action therefore extends only as far as I can execute my will through physical motion (25). No one has the power to touch the moon; and someone with a spinal injury cannot lift a cup.
Having dispatched freedom of action in his second lecture, Windelband moves on to freedom of choice, a much more complex and controversial topic, which takes up the next four lectures (lectures 3–6). The belief in freedom is pre-eminently a matter of the feeling that we have a choice, that we could choose equally between different courses of action, and that it is entirely up to me what I choose (33). Freedom of choice demands that I have the capacity not only to will but also to act on all options; if I am constrained so that I cannot do one or the other, or both, my freedom is limited accordingly (34). Nevertheless, Windelband observes, I can still choose one thing over another even if I cannot act on my choice; I can have preferences even if cannot act on them; it’s just that choosing loses its point or purpose—what Windelband calls its Zweckbedeutung—if I do not have the power to act on it (34).
Windelband distinguishes freedom of choice from moral freedom. Freedom of choice is a psychological concept; it refers to our unconstrained choosing and acting. It is a formal concept because it is irrelevant what we choose (93). The concept of freedom in a moral sense, however, is a normative concept; it concerns not what is but what ought to be (96). In this moral sense freedom has a negative meaning because it means independence from desire and passion; but it also has a positive sense because it signifies acting according to reason (95). These senses of freedom are plainly distinct from one another, Windelband insists, because a person can have freedom of choice but not moral freedom (96). This is the case when he or she acts immorally with perfect resolve and through reflective choice. Though they are plainly distinct, these senses of freedom are confused constantly, Windelband complains. We call a person unfree if he acts on base passions, and free if he acts on higher ones. But in both cases we are equally free in terms of freedom of choice. Windelband’s complaints about this confusion applied to no one more than himself, because, as we have seen,35 he was guilty of it in ‘Normen und Naturgesetze’.
The question of freedom of choice raises the hoary conundrum of whether there can be acts of choice without causes, motives or reasons? Windelband admits the possibility of a choice between equal options when there is no sufficient reason for choosing one over another (42). He gives this example: one comes to a round surface of grass in the middle of one’s path; one can go around it by moving right or left; there is no sufficient reason for going one way or the other.
In this sense it is necessary to admit “the reality of a liberum arbitrium indifferentiae” whose basis lies in a “so to speak principium rationis deficientis” (42). But if Windelband accepts the possibility of a choice where there are no sufficient reasons for one option or another, he denies that there is such a possibility in the sense of an act of will for which there is no cause. He distinguishes between reasons and causes, so that even if there is no reason for moving right or left, or choosing one lottery number over another, there is still a cause for such an action (52). The principle of sufficient reason holds without exception if by “reason” one means simply a cause. Even in the cases where we are not self-conscious of the cause, there is still one there. In cases like moving to right or left or the lottery number, we allow something within ourselves to make the decision for us, namely, our motor mechanism, which has nothing to do with motives of will (45, 47).
Regarding the causes rather than just reasons for actions, Windelband is happy to affirm the central thesis of the determinist, that is, that there are causes for all human actions and decisions, such that if the causes are present, the action or decision happens of necessity (75). All philosophers are determinists, he asserts somewhat boldly, insofar as they hold what we will and do is the necessary product of our momentary and constant motives, where these constant motives make up what we call “character”. Such determination of action through character Windelband calls self-determination (Selbstbestimmung) (76). When we say that we are the cause of our own actions, we mean that we are self-determining in just this sense, that is, that the actions follow from our character.
Freedom of choice (Wahlfreiheit), Windelband goes on to say, just means that we choose to act according to our characters (76). If this is so, he argues, then the whole dispute between determinist and indeterminist is in the end a verbal matter, because all the indeterminist wants to say is that the will itself, in addition to all other motives, gives one motive (i.e. cause) for the action; he wants us to recognize something constant in our character as at least one of the motives or causes for the action. The indeterminist is opposed to determinism only because he thinks that all motives have to be external to character. But once we recognize that motives can also be internal to the character, belonging to his or her very nature or self, then the quarrel between determinist and indeterminist vanishes (78).
In attempting to resolve the dispute between determinism and indeterminism in this manner, Windelband adopts the classical compatibilist position, according to which freedom is possible within a completely determined universe. This was the position that he had already advanced in ‘Normen und Naturgesetze’, so it seems as if he has done little more than arrived, through a more circuitous route, where he already stood. It seems as if Windelband thinks the whole dispute about freedom has already been settled simply by adopting his compatibilism. His lectures might as well have ended, it seems, with the sixth lecture where he resolves the dispute between determinist and indeterminist.
But Windelband does not leave matters here. The whole issue of freedom becomes much more complicated for him when he comes to the third concept of freedom, freedom of the will. Now Windelband seems to put on hold his earlier compatibilism and to consider a more refined and qualified position. His analysis becomes more intricate and involved, partly because he now breaches the issue of responsibility, and partly because he enters the field of metaphysics. Windelband now distinguishes the three concepts of freedom along the following lines. While freedom of action is a psychophysical concept, and while freedom of choice is a psychological one, freedom of will is a metaphysical concept (123). Into the dark woods of metaphysics Windelband now enters. And the deeper into it he goes, the more murky his account becomes.
The metaphysical aspect of the problem of freedom begins for Windelband with the idea of first causes or causeless willing (123). We find ourselves presupposing this idea, he says, because of the concept of responsibility, which seems to demand that there are causes that are not the effects of other causes (122). This idea is problematic, however, because it appears incompatible with the principle of causality, which holds for all events in nature, whether psychological or physical (125). To resolve this apparent problem, Windelband advises caution, and more specifically noting different senses of causality. It is necessary to distinguish, he argues, between two senses of causality: that which holds between different events, and that which holds between a substance and individual events (126). When we talk about the determinism of nature we are referring to the first sense; but when we speak of metaphysical freedom we mean the second. While first causes or causeless willing is incompatible with the first sense, it is still compatible with the second. The principle of causality that holds sway over nature holds only for events, and it does not pretend to hold for some substance that is not within the series of events. For this substance, Windelband says, we can apply the concept of a first cause or causeless willing (127).
The metaphysical concept of freedom makes sense, Windelband explains, only when we apply it to the whole personality, not to individual acts of will and the events that are their effects (154). A person himself with his character stands outside all the series of events, even if all his individual acts fall inside it (155). The person or character as a substance is contingent, having no necessary causes, whereas the occurrence of all events is necessary (155). Windelband says that we can explain this concept of personality with the scholastic concept of aseïtas, that is, existing for onself as opposed to existing from or because of others (155). The concept of aeseity is essentially that of self-determination, where the self-determining is that which determines itself into action alone and is not determined into action by anything else (158). Spinoza had formulated this concept with his causa sui and Fichte with his self-positing ego.
Having taken his reader deep into the metaphysical woods in his tenth lecture, Windelband then struggles to take him out of them in his final two lectures. The concept of metaphysical freedom, which he seemed to endorse, is now suspended and rendered superfluous. We are now reminded that if the will were really without a cause, then we could not hold someone responsible (199–200, 210). Attributions of responsibility imply that we can act on someone’s feelings and dispositions (through admonition or punishment) to make them act differently, so that it ultimately presupposes the principle of causality (210). “The whole procedure of making someone responsible is completely bound up with, both backwards and forwards, knowledge of empirical causality.” (211) When we hold a particular person responsible, all we have to presuppose is the concepts of freedom of action and freedom of choice; but we do not have to presuppose that the person has some metaphysical power of will, that he or she had the power to act countercausally, so that they could have done otherwise on this particular occasion. To say that a person could have done otherwise really means that he or she should have done so, or that they would have done so if their character had been different and the way it should have been (212).
Windelband attempts to remove the metaphysical connotations of moral responsibility by talking about two ways of considering (Betrachtungsweise) human actions (195). We can consider human actions according to norms, where we appraise their value or worth; and we can consider them under laws, where we explain their occurrence. When we consider these actions under norms, we abstract from their causal explanation and focus on their compliance with rules, customs or laws. The concept of freedom of the will, of a causeless willing, is simply the result of abstracting from the causal explanation of the action; it is a way of saying that treating the causes of the action is irrelevant to its moral appraisal; but it does not mean that the action really has no causes, or even worse that it has mysterious causes that operate only by interrupting the continuity of nature (197). When we consider someone responsible for an action that he or she should not have done, this does not mean that we have some retrospective metaphysical knowledge of the action but only that we are justified in judging it according to norms, and that we are going to do something to the person so that he or she does not commit the offence again (218).
So, in the end, Windelband interprets the normative and the natural not as different ontological domains but simply as different methods or perspectives for regarding human actions. The difference between the normative and natural is not that between kinds of entity—something noumenal and phenomenal—but between different discourses about human actions. Just as we abstract from the causes of a belief when we judge its truth, so we abstract from the causes of action when we ponder its morality. While everything that exists falls under the natural perspective, it is not the only valid perspective on human action, the only method of dealing with it (195). There is not only the realm of existence but also the realm of norms, which have a sui generis logical status that does not imply the existence of anything. The problem with the metaphysical conception of freedom is that it treats the normative dimension as if it were somehow ontological, implying the existence of a distinct kind of will and soul. But once we grasp the sui generis discourse of norms, we realize that we are not bound to special ontological assumptions. We can proceed to evaluate human actions and assess responsibility with perfect legitimacy without metaphysics.
The origins and development of Windelband’s normative conception of philosophy is a complicated and, as we shall soon see, controversial tale. Still, it is worthwhile to tell it because it alone fully explains the purpose and meaning of his mature philosophy. Windelband did not always define philosophy in exclusively normative terms, and he came to his later views about philosophy only by gradually working his way through earlier ones. We need to know why he kept, and why he discarded, elements of his earlier views to understand the rationale behind his later ones.
Windelband’s first foray into the territory of epistemology was his habilitation thesis, his Über die Gewißheit der Erkenntnis, which first appeared in 1873.36 The central purpose of this little tract is to answer some big questions: “What is certainty?”, “How is it possible?” and “By what means is it attainable?” (16) Windelband raises these questions because he thinks they concern all of us as moral and practical beings. We all have “the thirst for certainty” because we fear that our most important beliefs—those by which we lead our lives—could be false. The sceptic arouses this fear in us because he brings all beliefs into question. We therefore turn to philosophy, which is the only discipline to address these doubts and to offer us deliverance from them.
In raising the classical problem of certainty, and in posing the dangers of scepticism, Windelband was deliberately, but cautiously, nudging philosophy towards foundationalism again. He defines philosophy as “the investigation into the ultimate grounds of being and thinking” (15). He then declares explicitly—as if he were Descartes or Reinhold—that there is still “the deep need for a deductive foundation of knowledge” (17). Philosophy has to proceed in this direction, Windelband assumes, because only such a foundation will slake the thirst for certainty and respond to the sceptic’s doubts. Windelband is also taking issue with those of his contemporaries who maintain that philosophy should be nothing more than an investigation into the logic of the sciences. In his view, there are two problems with such a conception of philosophy: it accepts the methods and results of the empirical sciences as if they were dogmas; and it ignores the more general philosophical questions central to all cultural life (17).
Still, though foundationalism has its merits, Windelband makes it clear that he is not calling for a rehabilitation of the speculative idealist tradition. Although the great idealists were correct to address basic questions, they went too far in attempting to answer them entirely on an a priori basis. The time is over, he declares, when philosophy can chase “the phantom of a completely creative knowledge in the blue air of the imagination” (18). “The wax on the wings of Icarus has melted”, so that philosophy has crashed to earth where it must learn to dwell among “the realm of facts”. Philosophy has to orient itself around the results of the exact natural sciences, and it has to walk hand-in-hand with them.
Having made this concession, Windelband goes no further. We are told in no uncertain terms that there are different ways for philosophy to collaborate with the sciences, and that it must never let down its critical guard by simply accepting the methods, presuppositions and results of the sciences as if they were the only possible. The positivists are guilty of making a dogma out of the sciences, as if they stood above all criticism. Without mentioning names, Windelband takes issue with the likes of Helmholtz and Lange, who too readily place their trust in the empirical sciences, especially the latest psychophysical research. That research, he does not doubt, has confirmed one of the main results of the critical philosophy: that the content of perception is the result of our perceptual and intellectual functioning. But the advocates of such an approach, Windelband protests, have not taken into account its implications for natural science itself. Namely, if all the content of our perceptual world is created by the mind, how can we claim objective truth for science itself? The results of psychophysical research have posed yet again the question of the foundation of the sciences, and it is necessary for philosophy to deal with it (20).
Windelband was not to remain entirely true to the foundationalist conception of philosophy outlined in Über die Gewißheit. Never would he attempt to provide anything like a deductive foundation of the sciences along the lines of Reinhold and Fichte. In ‘Was ist Philosophie?’ he is clear that we cannot prove the basic principles of logic or ethics without presupposing them. But, to an extent, Windelband never lost his foundationalist concerns, insofar as he always insisted, until his very last breath, that philosophy should determine and assess the basic principles of knowledge, morality and art.
The central and characteristic feature of Windelband’s early conception of philosophy, in contrast to his later conception, is its syncretism, its holistic approach to epistemology. In Über die Gewißheit Windelband stresses that philosophy should be not only a logical but also a psychological and metaphysical enterprise (32). This inclusion of psychology and metaphysics within the domain of epistemology is in marked contrast to his later strictly logical conception. Windelband not only preaches but practises this holistic approach. The account of certainty in Über die Gewißheit is pursued in not only logical but also psychological and metaphysical terms. We bring all these disciplines together, Windelband proposes, when we show “how a psychological process, by means of logical functions, is able to grasp metaphysical truths” (32). And with that as his goal, Windelband tries to show the interweaving of these logical, psychological and metaphysical factors.
In sharp contrast to his later conception of epistemology, the young Windelband is very emphatic about the need for psychology in epistemology. Any epistemological account of knowledge ultimately presupposes a psychological one, he argues in Über die Gewißheit, for the simple reason that thinking and knowing are mental activities. The limits, worth and justification of knowledge has to be determined by its purposes, which are the result of some psychological process (22). Although we can abstract the forms of judgement and reasoning from the psychic activities that create them, we need psychology to grasp “the actual driving mechanism behind knowing” (23). Formal logic tells us how we think; but it cannot tell us why or that we think, and even less what we think (24). It cannot even tell us why (i.e. the causes or mechanism) we draw the inference S=P from the premises M=P and S=M (23). The motive and power to do that comes from our psyche, about which logic tells us nothing.
As much as the young Windelband insists on the importance of psychology, he is very far from reducing epistemology down to psychology alone. The logical aspect of epistemology is of no less importance to him. Logic and psychology are for him two very different yet complementary approaches to knowledge. Windelband explains the difference between logic and psychology through two different senses of necessity. There is a causal or psychological sense of necessity, according to which a process of thinking is necessary because it conforms to natural laws. In this sense all thinking, whether valid or invalid, is necessary, simply because it is a natural event and all events conform to laws. But this is not the sense of necessity we have in mind in logic, because in logic we are concerned not with all thinking but solely with valid thinking (82). While both correct and incorrect thinking obey psychological laws, only correct thinking complies with logical ones. The kind of necessity involved in logical laws, which is more akin to the laws of ethics than psychology, is that of an imperative: they tell us what we ought to do, or how we should proceed, if we are to think correctly (85). We can make these logical laws into our habits, so that we learn to comply with them in all our thinking, and so that they become firm psychological realities; but we must not let this fact confuse us, so that we think that logical laws are only psychological in meaning and purpose (86).
In Über die Gewißheit Windelband provides an explanation of logical necessity that sheds much light on his later concept of normativity. The laws of logic turn out to be hypothetical imperatives: they prescribe means to ends. They tell us what we ought to do if we are to achieve a certain end: namely, the acquisition of knowledge. The laws of logic therefore lay down the necessary formal conditions for the acquisition of knowledge. They are, as Windelband calls them, “laws of ends, norms” (Zweckgesetze, Normen). This is the first instance in which Windelband uses the redolent term “norms”, which will later become a mantra for him and the Southwestern neo-Kantians. We can see now why Windelband talks about a “teleological” justification of norms: they are the means by which we attain ends. If they prove to be effective means for these ends—if they are sufficient and necessary to achieve them—then they we have all the justification we could give for them.
Only at the close of Über die Gewißheit does Windelband introduce the third dimension of his epistemology: the metaphysical. As a student of Lotze, Windelband was fully convinced of the importance and necessity of metaphysics for epistemology, and he reaffirms his teacher’s thesis that the problems of epistemology are ultimately resolvable only through metaphysics. “The question of the essence and possibility of knowledge conceals within itself,” he writes in his foreword, “the whole puzzle of existence”. For Windelband, it is the psychological dimension of epistemology that leads straight to the metaphysical, for that dimension regards knowing as an event, as a coming-into-being, as Aristotle would say, and to understand an event we need metaphysics (119). Herbart had always said that his psychology presupposes metaphysics, and Windelband heartily agrees, making it the foundation of psychology.
Windelband gives another reason for adding metaphysics to epistemology. The task of epistemology is to explain the relationship between the subject and object. But its analysis of that relationship always bumps up against something ultimately given—the particular content of sensation—that cannot be derived from the laws of our subjectivity. What this given is, and in what relations it stands to experience and other things, is the concern of metaphysics (105). Here again Windelband was gesturing towards Herbart, who made the ultimate simples of experience into the object of metaphysics. He expressly approves Herbart’s view that metaphysics culminates in the theory of knowledge, because that theory treats the highest form of being, which is knowledge (119). On this account, the theory of knowledge virtually became a special form of metaphysics.
Already by the early 1870s, then, Windelband had worked out a conception of philosophy that was deliberately syncretic, combining logic, psychology and metaphysics. There was a normative element in his early conception of philosophy, because he explicitly conceived of the disciplines of logic and ethics in normative terms. But this normative element was only one part of a much broader epistemological programme, which included psychology and metaphysics as integral components. Before he could arrive at his later conception of philosophy, then, he had to abandon his earlier syncretism, dropping psychology and metaphysics from his epistemological programme.
The first sign of a shift in Windelband’s syncretic position appears in a review he published in 1874 of Christoph Sigwart’s Logik.37 Here Windelband shows himself to be sympathetic to Sigwart’s project for a purely formal logic. The question for him now is how logic can be autonomous, a discipline in its own right, independent of psychology and metaphysics. Psychology and metaphysics are now presented as the Scylla and Charybdis that logic should avoid if it is to be an autonomous discipline. On the one hand, logic has to avoid conflation with psychology, which gives no place to the concept of truth. From a purely psychological standpoint there is neither truth nor error because all thinking, whether true or false, conforms to natural laws. The distinguishing feature of logic from psychology is precisely that it retains the concept of truth and shows the means to attain it. On the other hand, logic should also steer clear of metaphysics, which poses too many unsolvable problems. As soon as we raise the epistemological question of the meaning of truth, we are caught in metaphysics; hence logic should avoid epistemology as much as psychology. The key to a purely formal logic, Windelband explains, is a strictly normative conception of truth, according to which truth consists in normative necessity, that is, in the need to follow norms to attain the ends of knowledge. This normative necessity, which is about what we ought to think, is very different from the natural or psychological necessity, which is about what we must think. If logic is understood as such a normative discipline, Windelband argues, then we steer clear of psychological questions about how we must think as well as metaphysical questions about whether logic conforms to being itself. The question will be only whether thinking conforms to its own immanent laws, whether it is consistent with itself. Whether it corresponds to a higher metaphysical truth can be left aside as unnecessary to determine the validity of thinking in particular cases.
The review of Sigwart’s Logik shows Windelband concerned to separate rather than unify the disciplines involved in his syncretic project. It was significant that Windelband saw the normative dimension of logic as the means to separate it from psychology and metaphysics: he will later make the same point to provide an autonomous conception of philosophy. Still, it would be a mistake to assume that the syncretic project has fallen apart, that Windelband now thinks that logic, psychology and metaphysics should be separated. After all, Windelband is concerned in the Sigwart review only with formal logic, which is only one part of epistemology. While logic should separate itself from psychology and metaphysics, the same does not hold for epistemology, a more general discipline which attempts to unite logic, metaphysics and psychology. Indeed, even in the Sigwart review Windelband stresses that formal logic should not completely isolate itself from epistemology and psychology, because it is only by constant reference to them that it will be able to show how logic is applicable and relevant to attaining knowledge of the actual world.
That Windelband was far from relinquishing his syncretic project becomes clear from an article he published a year after the Sigwart review, ‘Die Erkenntnislehre unter dem völkerpsychologischen Gesichtspunkte’, which appeared in 1875 in Zeitschrift für Völkerpsychologie und Sprachwissenchaft,38 the very organ in which Cohen made his intellectual debut a few years earlier. This article very much reaffirms the syncretic approach of Über die Gewißheit. Following the standpoint of its editors, Windelband’s article endorses the value of having a psychological, and indeed anthropological or “historical-cultural”, perspective in epistemology. Although the laws of logic and ethics appear eternal and immutable, one must not assume, he argues, that they exist in some self-sufficient noumenal realm, as if they could be isolated from social and historical change. These laws too are the product of history and culture, and we become aware of them only through education into a cultural tradition and only through the growth of language and linguistic sensibility (174). It is also a serious mistake, Windelband adds, to assume that the forms of logic have no reference to human psychology. The very opposite is the case: these forms have “sense and meaning” (Sinn und Bedeutung) only with reference to psychic activities (167). They are norms that make sense only if they direct and regulate the activity of thinking, so that if there were no thinking they would completely lose their point and purpose. All the basic laws of logic presuppose the ends of discourse (viz. persuasion, correction and criticism), the basic human activities of communication. Even the law of contradiction rests on a psychological foundation, because it makes sense only in the context of affirming and denying, which are speech acts (170). The principle of sufficient reason too presupposes a psychological activity, because its demand for reasons serves the purpose of human persuasion, which is trying to get someone to have the same thought processes as myself (171).
The psychologism and historicism of this early article are clear and emphatic, so much so that it has been seen as the work of “a relativist and pragmatist of the first rank”.39 But it is important to see that Windelband never departs from his earlier belief in the universality and necessity of logical norms. He closes his article by reassuring the logician that explaining the genesis of logical laws is perfectly compatible with their universal and necessary validity. “The dignity of logical laws as absolute authoritative norms remains entirely intact.” (177) Still, there can be no doubt that, in this early article, Windelband is bent on avoiding a strictly logical or normative approach to epistemology, and intent on complementing it with psychology, anthropology and even cultural history. Nothing, it seems, could lead Windelband away from his syncretic conception of epistemology. But, little did he know, there were strong centrifugal forces lying just around the corner.
The first cracks in the syncretic project appear shortly after Windelband’s article in the Zeitschrift für Völkerpsychologie. On May 20, 1876, he gave his inaugural lecture as ordinary professor in Zurich, Über den gegenwärtigen Stand der psychologischen Forschung.40 Windelband chose as the topic for his lecture the question of the relationship between philosophy and the empirical sciences. The topic seemed especially appropriate for his new position, because his chair was devoted to “inductive philosophy”, an attempt to modernize philosophy by aligning it with the new empirical sciences. True to that spirit, Windelband reaffirmed his position in Über die Gewißheit about the importance of close collaboration between the empirical sciences and philosophy. The bad old days, when philosophers believed that they could achieve meaningful results simply through a priori speculation, were long over and should never return.
Having paid his homage to the new sciences, Windelband then turns around and insists that philosophy has to guard its borders from them. Philosophy and the empirical sciences, we are now told, need to respect their proper boundaries, and this is especially the case for philosophy and psychology (3–4). If these disciplines are to collaborate, each must know its special task, so that it is more effectively pursued. Windelband was thus reasserting the very academic division of labour that his syncretic approach was designed to overcome.
In stressing the importance of proper boundaries between philosophy and psychology, Windelband was addressing an issue that especially concerned German academic philosophy in the late 19th century: that because of the porous border between philosophy and psychology, positions once intended for philosophers were becoming increasingly filled by psychologists. This was the ultimate humiliating outcome of the identity crisis. If philosophers did not know themselves and set firm boundaries, they were going to be replaced by psychologists.
Windelband’s strategy in responding to the crisis is the very opposite of what we might first expect. Rather than declaring the independence of philosophy from psychology, he recommends the independence of psychology from philosophy. Windelband chose this strategy, partly because it was more appropriate for a chair in “inductive philosophy”, and partly because it was more effective in achieving his ultimate end: autonomy for philosophy. If you advocate the autonomy of psychology, psychologists will cease to compete with you; instead, they will lobby for their own positions independent of philosophy, which is what philosophers want too.
Windelband’s argument for a sharp boundary line is simple and straightforward. Psychology will make progress, and be assurred of achieving verifiable results, only to the extent that it becomes completely independent of metaphysics. Psychology must free itself from metaphysics for two reasons: first, because co-operation requires agreeing about the basics, and no one agrees about the basics in metaphysics; and, second, because metaphysics has all too often resorted to hypothetical explanations and abstract concepts (viz. the soul and Lebenskraft) which impede further empirical enquiry (6–10).
Assuming that philosophy should leave psychology to its own pursuits, what is left for it to do? Windelband only suggests his answer at the very end of his lecture. He now conceives the task of philosophy chiefly in logical terms: “the justification of the methods of scientific research and the foundation of the principle forms of conceiving and explaining” (24).
The implications of this lecture for Windelband’s conception of philosophy are plain enough: the syncretic conception explodes, falling into three distinct pieces, logic, psychology and metaphysics. Honouring the academic division of labour means these should be separate disciplines, each pursuing their distinct concerns. Psychology should be separate from metaphysics, which violates its autonomy; and philosophy should be logic alone, leaving psychology to pursue its own empirical enquiries. Windelband did not draw all these implications immediately or explicitly; but their latent presence made them cracks in the wall of his fragile syncretic project.
That project came under further stress in an article Windelband wrote in the following year, his ‘Ueber die verschiedenen Phasen der kantischen Lehre vom Ding-an-sich’, which appeared in 1877 in the second issue of the Vierteljahrschrift für wissenschaftliche Philosophie.41 This article is primarily historical in content, dealing with Kant’s intellectual development, and more specifically the evolution of his doctrine of the thing-in-itself. One might think, therefore, that it holds little of interest for Windelband’s philosophy. But, given Windelband’s growing identification with Kant’s philosophy, its implications for his own position prove to be considerable. We need not consider here all the details of Windelband’s interpretation. Suffice it to say that he thinks that Kant, during his critical phase, held three incompatible views about the thing-in-itself: 1) that the thing-in-itself is something unthinkable and therefore impossible; 2) that it is something that we can assume even though we cannot know it; and 3) that it is something that we must assume to explain the world of appearances (256–257). Anticipating the patchwork theory, Windelband thinks that Kant held these views at different times, that they were written down in distinct fragments, which he then stitched together in the final months of the composition of the Kritik der reinen Vernunft (231–232). We can now see why Windelband famously said understanding Kant means going beyond him. We simply cannot, like Kant, hold all these views; consistency alone demands moving beyond the confusions inherent in the texts. One view alone can represent the spirit of the critical philosophy.
Which view should that be? Windelband is under no doubt about it. He thinks that the first view represents Kant’s deepest intentions. The proper conclusion to draw from Kant’s transcendental deduction, and the limits he places on knowledge, he argues, is that the thing-in-itself is a complete impossibility. The central thesis of the transcendental deduction is that objectivity derives not from the correspondence of representations to some entity that exists independent of them, but from the conformity of sensations to the rules of synthesis of the understanding. On this reading, the idea of a thing-in-itself is simply the hypostasis of the concept of an object, which really derives from the synthesizing function of the understanding (254). And the limits Kant places on knowledge—possible experience—mean that it is not possible to assume the existence of anything that exists beyond it (254). For these reasons, Windelband comes to the conclusion that the real spirit of Kant’s critical idealism lies in the abolition of the thing-in-itself, which is only a hangover from a naive realistic view of the world that Kant intended to overthrow. And so Windelband declares his death sentence: “From the standpoint of epistemology, the thing-in-itself is a completely meaningless and useless, confusing and annoying, fiction” (258). The spirit of the critical philosophy resides in its principle, which is presupposed but never fully expressed, that knowledge must be explained completely within itself and not by transcendent postulates about something outside itself (260). “The divine gift of Kant to humanity,” Windelband adds, lies in “this thought that there is nothing beyond representations” (261). He admits that there is no particular text that unambiguously expresses the spirit of Kant’s philosophy. Kant comes closest to it in the chapter ‘On the Distinction between Phenomena and Noumena’ where he states that the very distinction between appearances and things-in-themselves is untenable; but even that text was altered in the second edition to make it accord with the hypothetical existence of things-in-themselves (248 n.10).
The implications of Windelband’s interpretation of Kant for his general conception of philosophy are immediate and striking. Since Windelband stresses Kant’s “critical idealism”, which limits all knowledge to experience, epistemology is now purged of metaphysics. The pivotal role Windelband was ready to assign to metaphysics in Über die Gewißheit vanishes. Perhaps there is “a secret of being” lying in the depths of epistemology, as he once said; but he now realizes that the epistemologist is in no position to tell it.
There is another crucial implication of Windelband’s interpretation, though it is less obvious. Towards the end of his article he insists upon connecting two aspects of Kant’s philosophy that are apparently distinct: the postulate of the thing-in-itself and his psychology. When we drop the former, Windelband argues, the latter too disappears. Why? The explanation is perfectly straightforward. Kant’s investigation into the possibility of knowledge originally arose from the question how a priori representations correspond to an object that exists independent of them. If these representations arise from the mind, if they are not the effect of the object itself, then how do they correspond to objects? This question assumes a thing-in-itself which exists independent of representations; and it is posed in psychological terms, because it deals with the causal relations between representations and their objects. If, however, we drop the thing-in-itself entirely, insisting upon explaining knowledge in entirely immanent terms, then the focus of enquiry is no longer on the causes but the content of knowledge. In other words, logical enquiry replaces psychological (259). The more Kant became involved in the argument of the transcendental deduction, Windelband argues, the more he could see that all the assumptions about the dualism between subject and object, and the causal relations between them, should not be the starting point but the subject matter of investigation. Seen properly, then, psychology cannot be the key to epistemology, as Fries, Beneke and Meyer preached, because it is really part of the problem. Kant himself, Windelband thinks, saw this point, even if through a glass darkly. It explains why, in the preface to the second edition of the Kritik, he demoted the importance of psychology, and why, in the Prolegomena, he formulated his problem in the terms of the possibility of synthetic a priori judgements. A purely immanent explanation of the possibility of knowledge thus became an essentially logical enquiry, an examination of the content of representation rather than its causes, of the quid juris? rather than the quid facti?
And so not only metaphysics but also psychology was ejected from Windelband’s conception of epistemology. Already by 1877, then, we find Windelband close to the strictly normative conception of philosophy that he will preach in the early 1880s. Windelband came to this conception chiefly from two considerations: first, the academic division of labour, which demanded separating philosophy from psychology and metaphysics; and, second, his interpretation of Kant’s philosophy as critical idealism. From these considerations alone, it seems that Windelband’s normative conception was utterly overdetermined, the necessary result of his thinking before 1878.
As plausible as all this sounds, not everyone agrees with it. One scholar finds Windelband’s crucial turning point after 1878; and, worse still, he sees all our reasoning as suspect. We must now give him a fair hearing.
According to Klaus Christian Köhnke, the account we have given so far of Windelband’s intellectual development is naive.42 It places Windelband in “a stage of innocence” that would explain his intellectual development entirely from systematic and philosophical considerations, as if he were a pure intellect who moved entirely in aetherial orbs. This is naive, Köhnke implies, because it abstracts Windelband from his social and political context, underestimating the political motives behind the formation of his philosophical views. Köhnke is convinced that there is one political event in particular that was decisive for Windelband’s intellectual development, and that was the source of his normative conception of philosophy: the two assassination attempts on the life of Kaiser Wilhelm I in 1878.43 This aroused a hysterical reaction among conservative politicians and the bourgeoise, who abhorred the growing threats of socialism and democracy. This reaction eventually led to Bismarck’s introduction of an Anti-Socialist Bill in the Reichstag, which prohibited working class associations and meetings. Windelband, Köhnke believes, was among those “German mandarins” who feared working class agitation and universal suffrage,44 which were alarming developments that threatened his entire world, not only elite rule but German culture itself. Socialism and democracy meant “levelling”, the decline of moral, aesthetic and cultural standards to meet the demands of the ignorant masses. Windelband’s chief defence against these dangers, Köhnke maintains, was nothing less than his normative conception of philosophy. It was philosophers alone who could and should be the standard bearers of culture by representing its universal and necessary values. The concept of “normal consciousness”, as Köhnke puts it, was “a weapon (Kampbegriff) against democrats, republicans and socialists alike”.
Dwelling in our stage of innocence, we might well be astonished that a philosophy that pretends to preserve and promote universal and necessary values somehow serves as a bulwark for “the interests of the authoritarian Bismarckian state”.45 This seems unduly partisan and partial, given that universal values should represent everyone alike, all intelligent or rational beings, whether they be stout Bismarckians or rabid socialists. But, arguably, this too is naive, because so much depends on who knows, and who administers, these values. If the Windelbandian state is ruled by philosophers, or at least princes advised by philosophers, then knowing universal values demands intelligence, education and talent. The masses, who lack such merits, cannot know these principles, and so they should not govern; or, in more concrete political terms, they should not have the right to vote.
So Köhnke’s political account of Windelband’s development is more plausible than it might first seem. The case for his theory ultimately depends on the concrete historical evidence he marshals for it, more specifically, evidence concerning Windelband’s political views in the late 1870s and early 1880s. But before we examine that evidence, we need to consider Köhnke’s argument for resorting to a political explanation in the first place.
Ironically, Köhnke turns to Windelband’s politics and the events of the late 1870s because he finds a systematic or purely logical need to do so. There is, he argues, “a break” in Windelband’s intellectual development in the late 1870s, indeed “a conversion from Saul to Paul”.46 The break is away from a psychological-historical approach to logic, ethics and aesthetics to a normative one. In Über die Gewißheit and the article for the Zeitschrift der Völkerpsychologie Windelband is “a relativist of the first order”, Köhnke thinks, because he advocates a radical psychological and historical standpoint, according to which the norms of logic, ethics and aesthetics derive from psychology and history. But this relativist Saul became an absolutist Paul in the late 1870s. Why the sudden conversion? Politics, it seems, is the only plausible answer. This alone explains the shift from a champion of psychologism and historicism to an advocate of universal values and absolute norms.
Something can be said on behalf of Köhnke’s account of Windelband’s intellectual development. We have seen that there was indeed a break in Windelband’s development in the 1870s, a move away from a syncretic conception of philosophy to a strictly normative one. But, apart from this very general point, Köhnke’s account of this break is inaccurate, on both historical and philosophical grounds. Historically, the break took place not in 1878, the year of the assassination attempts, but earlier, in 1876 with Windelband’s inaugural lecture in Zurich, and in 1877 with his article on the Kantian thing-in-itself. Philosophically, Windelband’s break was not a black-or-white affair, a move away from a completely historical and psychologistic conception of epistemology to a totally normative one. Windelband’s normative conception of philosophy was already fully in place in the early 1870s in Über die Gewißheit; it’s just that it was also mixed with psychological and metaphysical elements. Never was Windelband the radical historicist and psychologist, the total relativist, that Köhnke makes him out to be. We have already seen how, at the end of the article in the Zeitschrift für Völkerpsychologie, Windelband reaffirmed the absolute status of logical and ethical truths. Behind Köhnke’s entire account there lingers a false assumption. Namely, that Windelband’s normative and historical-psychological approaches to epistemology are contradictory. It is for this reason that he has to portray Windelband’s development in such black-or-white terms. But the assumption would have been rejected by Windelband, who always insisted that these approaches are complementary rather than contradictory. Because we have to distinguish between validity and fact, norms can have a universal and necessary validity even though they arise from particular and contingent psychological and historical causes.
Granted that Köhnke’s account of Windelband’s break is mistaken, it is still possible to rescue a weaker version of it. For even if the normative conception of philosophy was already present in the early 1870s, it was not entirely or exclusively normative, and that entirety and exclusivity might well have a political explanation. Perhaps there were political reasons that motivated Windelband to advance a purely or strictly normative conception of philosophy? We have seen that Windelband had systematic reasons for developing that conception: the sharp distinctions he had drawn between philosophy and psychology are sufficient to explain its genesis. But they do not exclude a political explanation as well. So perhaps there were indeed political motives after all?
At this point everything rests on the evidence Köhnke provides about Windelband’s political views in the late 1870s. Köhnke rests his case chiefly on two lectures that Windelband gave around 1878: ‘Über Friedrich Hölderlin und sein Geschick’ and ‘Über Sokrates’.47 Both these lectures show Windelband to be a staunch political conservative. In them his opposition to democracy, to a universal franchise, is especially apparent. In ‘Über Friedrich Hölderlin’ Windelband laments the specialization, fragmentation and atomism of modern civil society, which stands in sharp contrast to the unity of the ancient world, where each individual could understand and feel part of his entire culture. The paltry surrogate for this modern malady, Windelband says, is “dilettantism”, the attempt to recapture a sense of wholeness by knowing a little bit about everything. This dilettantism appears in the salons and the alleyways. If it is comic in the salons, it is tragic in the alleyways, because those who practise it there are the demagogues, who use scientific themes to preach to the people. The political form of this dilletantism, Windelband then declares in a revealing passage, is “Parliamentarianism”, which would give everyone a right to speak about public matters (256). In ‘Über Sokrates’ the same contempt for democracy emerges. Windelband paints a picture of Socrates as a spokesman for the authority of reason against the shallow enlightenment of the sophists, who popularized the democratic idea that everyone should express their opinion about everything. That idea was dangerous because it undermined cultural unity and the authority of the laws. “There follows from the democratization of knowledge”, Windelband tells us in no uncertain terms, “the demoralization of culture.” (61) Windelband’s Socrates was a “Wertwissenschaftler” avant la lettre because he recognizes that there is a standpoint of a universal reason that stands above the individual and to which his judgement must submit (67–68). That standpoint thus serves as a bulwark against a rampant relativism and democratization, which would level all opinions and values.
So, much as Köhnke says, these lectures show Windelband to be a political conservative, and indeed during the social hysteria of 1878. But the crucial questions remain whether Windelband’s political beliefs have the meaning Köhnke reads into them, and whether as such they were the chief motive for his normative conception of philosophy. And here it is necessary to be sceptical. Köhnke tells us that Socrates’ standpoint of reason, as portrayed by Windelband in ‘Über Sokrates’, was meant to be “utterly authoritarian” (durch und durch autoritär gemeint).48 But Windelband stresses that Socrates’ standpoint is accessible to everyone alike, that it has to be confirmed by the judgement of each individual, and that it will appear as an external authority only for that individual who has not sufficiently reflected and engaged in a full enquiry. Köhnke ignores the fact that Windelband’s Socrates represents the standpoint of moral autonomy no less than rational authority, and that, like a good Kantian, he encourages every individual to think for himself. On Windelband’s interpretation, Socrates was not an opponent of enlightenment as such but only of a half-enlightenment which did not go far enough and stood only half way. There were three stages each individual went through, Windelband held, in his journey to philosophical enlightenment: blindly following external authority; liberation from that authority; and then rediscovering that authority from within through the exercise of one’s own reason (75). All this suggests that Windelband was not opposed to democracy or republicanism as such, but only to granting it to a public who were not ready for it because they lacked sufficient education—an all too common conservative belief in the 1870s and 1880s.
All the plausibility of Köhnke’s reading of ‘Über Sokrates’ rests on a notorious but mysterious fact: that Socrates preferred aristocracy as his ideal form of government. The precise role of this fact in Socrates’ trial has always been obscure. But to many it was decisive, making the trial into a contest between an aristocratic Socrates and a democratic Athens. This fact then makes it seem as if Socrates’ stand on behalf of reason was meant to be aristocratic, indeed authoritarian. But it is striking and telling that Windelband does not present Socrates’ trial in this light. It was a mistake of Socrates’ judges, he tells us, that they read his contempt for their tribunal as a contempt for democracy itself (84). The trial was not a contest between aristocracy and democracy but between individual autonomy and tradition (75–76). It is significant, Windelband points out, that Aristophanes, Socrates’ most passionate critic, never gave any importance to Socrates’ aristocratic beliefs (79–80). For Aristophanes, the issue was the social and political dangers of giving each individual the right to question the laws and traditions. Where Aristophanes misunderstood Socrates was in lumping him together with the sophists, whose shallow half-enlightenment he too regarded as dangerous.
The upshot of this closer reading of the text should be clear: that if Windelband intended Socrates’ standpoint of reason to be “a weapon” against all democratic tendencies, it was not very effective at all. For this standpoint was not meant to be esoteric, the preserve of philosophers alone. All were supposed to see its evidence and authority, if they were only properly prepared and educated. The point at issue then becomes whether Windelband was liberal enough to advocate equal opportunity for all to have a general education. And here there can be no doubt. In ‘Über Friedrich Hölderlin’ Windelband expressly advocates an education “for all classes and individuals” which ensures “equality in [its] intellectual foundations” (257).
The ultimate irony is that in diagnosing the problems of modern education and culture Windelband proposes moving philosophy in a direction opposed to his purely normative conception of philosophy. For at the close of ‘Über Friedrich Hölderlin’ Windelband laments the fact that there is no longer a general philosophy that can forge all the results of the sciences into a single worldview (259). Only such a philosophy would help to overcome the fragmentation and specialization of modern life. This was the true student of Lotze speaking, whose Mikrokosmus was an attempt to provide just such a worldview. But this demand for a worldview would eventually push Windelband away from the exclusively normative conception of philosophy and back towards his more syncretic approach.
Now, at the end of our foray into Windelband’s politics, we can perhaps rest content that our original explanation was not so naive after all. Köhnke’s political account does not begin to explain the nuances, details and complexities of Windelband’s philosophical development. To do that, we need to resort to systematic and internal philosophical considerations. The attempt to determine the rationale for a philosophical position through politics alone will never take us very far. The German mandarins, who might all have shared the same political interests, still held the most diverse philosophical positions.49
1 Windelband’s best-known historical work is his Lehrbuch der Geschichte der Philosophie (Tübingen: Mohr, 1900). It has been through no less than nineteen printings since 1900 and was published as recently as 1995. His chief works on the history of philosophy are his Die Geschichte der neueren Philosophie in ihren Zusammenhange mit dem allgemeinen Kultur und den besonderen Wissenschaften (Leipzig: Breitkopf & Härtel, 1878–80) and his Geschichte der alten Philosophie (Munich: Beck, 1894). Both works went through many editions.
2 In 1894 Windelband supported Julius Bergmann and opposed Cohen and Natorp in the habilitation case of Ludwig Busse, and he did so by declaring Bergmann alone a competent judge of the candidate’s qualities. In the early 1900s Windelband saw no place for Cassirer in Straßburg “on confessional grounds”; whether he opposed Cassirer personally on these grounds, or was simply reporting a fact about the administration, is difficult to determine from the context. On these facts, see Ulrich Sieg, Aufstieg und Niedergang des Marburger Neukantianismus (Würzburg: Königshausen & Neumann, 1994), pp. 192, 335.
3 Wilhelm Windelband, ‘Die philosophische Richtungen der Gegenwart’, in Große Denker, ed. Ernst von Aster (Leipzig: Quelle & Meyer, 1911), II, 361–377, esp. 370–371.
4 Hermann Cohen, System der Philosophie. Erster Teil: Logik der reinen Erkenntnis. (Berlin: Cassirer, 1902). System der Philosophie. Zweiter Teil: Ethik des reinen Willens. (Berlin: Cassirer, 1904). System der Philosophie. Dritter Teil: Ästhetik des reinen Gefühls (Berlin: Cassirer, 1912).
5 Wilhelm Windelband, Präludien. Aufsätze und Reden zur Philosophie und ihrer Geschichte. Neunte Auflage. (Tübingen: Mohr, 1924), two volumes. All references to this work will be to this edition.
6 Windelband’s interest in the logic of history began with his famous Straßburger Rektoratsrede Geschichte und Naturwissenschaft (Strassburg: Heitz, 1894). It continued with Heinrich Rickert’s Die Grenzen der wissenschaftliche Begriffsbildung (Tübingen: Mohr, 1902) and Die Probleme der Geschichtsphilosophie, Dritte Auflage (Heidelberg: Winter, 1924). The same interest appeared in Emil Lask’s Fichtes Idealismus und die Geschichte (Tübingen: Mohr, 1902).
7 On the philosophy of value, or so-called Wertphilosophie, see Herbert Schnädelbach, Philosophy in Germany 1831–1933. (Cambridge: Cambridge University Press, 1984), pp. 161–191.
8 This is especially the case with regard to the so-called “Pittsburgh school”. See Chauncey Maher, The Philosophy of the Pittsburgh School (London: Taylor & Francis, 2012).
9 There are few sources on Windelband’s life. The main source is Heinrich Rickert’s Wilhelm Windelband, Zweite Auflage (Tübingen: Mohr, 1929).
10 See his assessment of Fischer in Kuno Fischer und sein Kant (Halle: Ehrhardt Karras, 1897), pp. 10–11; and Kuno Fischer: Gedächtnisrede (Heidelberg: Winter, 1907), pp. 24–25.
11 See Wilhelm Windelband, ‘Otto Liebmanns Philosophie’, in Kant-Studien XV (1910), iii–x.
12 See my The German Historicist Tradition (Oxford: Oxford University Press, 2011), pp. 365–392; and my ‘Historicism and neo-Kantianism’, Studies in History and Philosophy of Science 39 (2008), 554–564.
13 Wilhelm Windelband, ‘Immanuel Kant. Zur Säkularfeier seiner Philosophie’. Ein Vortrag, 1881. Präludien I, 112–146.
14 Kant, ‘Vorrede’ to Metaphysische Anfangsgründe der Naturwissenschaft, AA IV, 470.
15 See Kant to Marcus Herz, February 21, 1772, Briefwechsel, ed. Otto Schöndörffer (Hamburg: Meiner, 1972), pp. 99–106.
16 Here is the decisive passage in the ipsissima verba of the German original: “In der unendlichen Mannigfaltigkeit der Vorstellungsverbindungen gibt es solche, welche einer allgemeingiltigen [sic] Regel, einer Norm entsprechen. Wahrheit ist Normalität des Denkens.” (Präludien I, 138).
18 See his 1904 lecture ‘Nach hundert Jahre. Zu Kants hundertjährigen Todestage’, Präludien I, 149–150.
19 Windelband’s student, Richard Kroner, developed this interpretation in his Kants Weltanschauung (Tübingen: Mohr, 1914).
20 Rorty places Kant within the epistemological tradition held captive to the mirror of nature metaphor. See his Philosophy and the Mirror of Nature (Oxford: Blackwell, 1980), p. 12.
21 Windelband, Präludien I, 1–54.
22 Wilhelm Windelband, Die Philosophie im deutschen Geistesleben des XIX. Jahrhunderts. Fünf Vorlesungen. (Tübingen: Mohr, 1909), pp. 81–84.
23 Windelband would argue this point in his 1875 inaugural lecture in Zurich, which we will examine in Section 9, and in his 1894 ‘Strassburger Rektoratsrede’, ‘Geschichte und Naturwissenschaft’, Präludien II, 143.
24 Rickert and Weber made the distinction between value judgements (Werturteile) and reference to values (Wertbeziehung), which they used to distinguish between the cultural and natural sciences. On their distinction and its problems, see Beiser, The German Historicist Tradition, pp. 412–414, 532–533.
25 Windelband, ‘Normen und Naturgesetze’, Präludien II, 59–98.
26 KrV B579n.
28 Windelband, Präludien II, 99–135.
29 See Windelband, ‘Über die gegenwärtige Lage und Aufgabe der Philosophie’, Präludien II, 1–23, esp. 21.
30 See Windelband, 1910 essay ‘Kulturphilosophie und transzendentaler Idealismus’, Präludien II, 279–294.
31 See Wilhelm Windelband, Geschichtsphilosophie. Eine Kriegsvorlesung. Fragment aus dem Nachlaß. Kant-Studien Ergänzungsheft 38 (1916).
32 Windelband, ‘Die Erneuerung des Hegelianismus’, Präludien I, 273–289, esp. 281, 284.
33 On the neo-Hegelian movement, see Paul Honigsheim, ‘Zur Hegel Renaissance im Vorkriegsheidelberg’, Hegel-Studien II (1963), 291–301.
34 Wilhelm Windelband, Über Willensfreiheit. Zwölf Vorlesungen. Zweite unveränderte Auflage (Tübingen: Mohr, 1905). All references in parentheses are to this edition.
36 Wilhelm Windelband, Über die Gewißheit der Erkenntnis. Eine psychologisch-erkenntnistheoretische Studie (Berlin: F. Henschel, 1873). There is a new edition of this text produced by Adlibri Verlag, Hamburg, 2005. All page references are to this new, more accessible edition.
37 Wilhelm Windelband, ‘Zur Logik’, Philosophische Monatshefte 10 (1874), 33–42, 85–91, 103–110. Although this article appeared one year after the publication of Über die Gewißheit, it appeared nearly two years after its composition, because, as Windelband tells us in the ‘Vorwort’ to that work, the publication of the almost printable manuscript was delayed for “personal reasons” for “almost a year”.
38 Wilhelm Windelband, ‘Die Erkenntnislehre unter dem völkerpsychologischen Gesichtspunkte’, Zeitschrift für Völkerpsychologie und Sprachwissenschaft VIII (1875) 166–178.
39 Klaus Christian Köhnke, Entstehung und Aufstieg des Neukantianismus: Die deutsche Universitätsphilosophie zwischen Idealismus und Positivismus. (Frankfurt: Suhrkamp, 1986), p. 362. We will examine Köhnke’s interpretation of the young Windelband in detail in Section 10.
40 Wilhelm Windelband, Über den gegenwärtigen Stand der psychologischen Forschung. Rede zum Antritt der ordentlichen Professur der Philosophie an der Hochschule zu Zürich am XX. Mai MDCCCLXVI (Leipzig: Breitkopf & Härtel, 1876).
41 Wilhelm Windelband, ‘Ueber die verschiedenen Phasen der kantischen Lehre vom Ding-an-sich’, Vierteljahrschrift für wissenschaftliche Philosophie Jahrgang I, Heft II (1877), 224–266.
42 See Köhnke, Entstehung und Aufstieg, p. 421.
43 On the historical and political background, see James J. Sheehan, German Liberalism in the 19th Century (Chicago, IL: University of Chicago Press, 1978), pp. 182–184.
44 Though it nicely fits with his analysis, Köhnke does not refer to Fritz Ringer’s The Decline of the German Mandarins: The German Academic Community, 1890–1933 (Cambridge, MA: Harvard University Press, 1969). Ringer includes Windelband among his German mandarins. Unfortunately, he offers no evidence about his social and political views. It seems that to be a mandarin one only needs to have been an academic during these years.
45 Köhnke, Entstehung und Aufstieg, p. 427.
46 Köhnke, Entstehung und Aufstieg, p. 421.
47 ‘Über Friedrich Hölderlin und sein Geschick’, Präludien I, 230–259; and ‘Über Sokrates’, Präludien I, 55–87.
48 Köhnke, Entstehung und Aufstieg, p. 427.
49 I have to express my scepticism, therefore, about Ringer’s attempt “to derive the opinions of the German academic intelligentsia from its peculiar role in German society.” Ringer, The German Mandarins, p. 4. All that Ringer shows us is that many German academics shared a belief in the need for and value of elite rule. This is far from explaining their social and political views, let alone their epistemological and metaphysical ones, which were very diverse. Ringer’s summaries of the more philosophical views of his mandarins are so clumsy and crude that it is questionable he understands their meaning or purpose.