1. The Ambivalence between Taylor’s Narration of the Self and its analytical Framework

For more than 30 years now, Taylor’s work has pushed us to our intellectual limits. As straight-forward and corresponding to our everyday comprehension of life as it might seem, its theoretical framework is immense and, if we dare to look closely enough, sometimes rather confusing. Perhaps resolving that ambivalence between the easy intuitive accessibility of Taylor’s work and its anything-but-easy analytical accessibility is the most demanding of all tasks Taylor puts before us. If one reads his work naively, without attempting to analyze every little detail or follow all the historical tracks of the origin of what he calls ‘the self,’ one may well see nothing more than a complete affirmation of what one already believes to be. But that is certainly not all Taylor has to offer. Even if it’s tempting to read his work as some sort of explication that just enlightens us to ourselves, as a narrative of what we have always already known, this would be a simplification, a reduction that ignores the philosophy behind the great and fascinating story Taylor tells us about our selves.

The reason that I have started this paper by pointing out a gap between an intuitive access to Taylor’s work and an analytical one, between his narration of our everyday lives and the philosophical foundation of this narration, isn’t just the invitation, the request to look closer, beyond the mirror he apparently holds in front of us. Over and above this, I would like to show that this gap, this ambivalence, between what Plato would call the ‘anamnesis of our selves’ and the framework in which such a self-recognition takes place is itself inherent in Taylor’s work. As implicit and—as I tend to think—well-hidden as it is in most of his earlier books, in The Language Animal it becomes obvious. It manifests itself in a serious struggle to theoretically frame the becoming of the self without abstracting it from its being-in-the-world, from the embodied being ^{Footnote 1} that isn’t framed by anything other than its concrete existence, its radical immanence, its ‘cultural reality.’

Although this isn’t a new problem, it now appears in its fullest form. Before The Language Animal, one could simply follow Taylor’s narration without asking for its own condition of possibility, for the philosophical framework that frames the framing of the self as a being-in-the-world. ^{Footnote 2} Or one could, as Richard Rorty does, just state that the realism Taylor imposes with his seemingly frameless framing of the self as a being-in-the-world is nothing but a “trivial, uninteresting, and commonsensical one which says that all true beliefs are true because things are as they are.” ^{Footnote 3} I think neither is possible anymore. Now the question is: is the framework that has been lingering around for so many years in its implicitness capable of closing the gap? Or does it undermine Taylor’s narration of the self and, indeed, has it done so all along?

In his earlier books, especially in Sources of the Self, the Archimedean point has always been the constitutive goods ^{Footnote 4} that can be described as the necessity of affirmation as such and, insofar as the conditio sine qua non of the (always already) affirmative narration of the self. ^{Footnote 5} But besides the plausibility of the logical necessity of an ‘affirmative stance,’ the concept of ‘goodness’ to be able to affirm, to differentiate between what to affirm and what not to affirm, the meaning as well as the theoretical foundation of those constitutive goods remained strangely obscure. Are they just plausibility or is there more to them—necessity? To my mind, the answer to this daring question comes finally to light when we take a close look at The Language Animal and particularly at Taylor’s recourse to Gottlob Frege therein. Emplacing this philosophical (and, of course, mathematical) heavyweight, whose Platonic logic forged a whole new understanding of what we call ‘reality,’ he puts himself in an analytical tradition that one (including himself) may not have expected. Although Taylor defies what he calls “the mainstream of post-Fregean analytic philosophy,” ^{Footnote 6} which he counters focusing on Frege’s context principle with his own interpretation, the Fregean distinction between sense and reference ^{Footnote 7} by itself sheds a clearly Platonic light on his concept of ‘reality’ ^{Footnote 8} and within it on his Archimedean point of the constitutive goods.

So is Taylor a closeted Platonist? There are indeed already some hints in his Sources of the Self: Like Plato’s idea of the good, the constitutive goods do “more than just define the content of the moral theory.” Instead they are “a something the love of which empowers us to do and be good” ^{Footnote 9} —or to put it differently: the condition of the possibility to be good, the concept of ‘goodness’ itself, the reference of everything that is good, the reference of every affirmation. Nevertheless, Taylor can’t be called a Platonic philosopher in the traditional sense. The title would not fit with his narration of the self as a being-in-the-world. It wouldn’t add up with the self not being framed by anything other than its concrete existence, its ‘cultural reality.’ Because of this, because it just wouldn’t add up with the rest of Taylor’s philosophy—at least not if only understood as some sort of explication that just enlightens us to ourselves, as a narrative of what we have always already known—the implicit (but obvious) Platonism in his work is often rejected. To secure his theory against the charge of incoherence, an obscure new ‘realistic position between Platonism and Projectivism’ ^{Footnote 10} has even been invented that is thought to fit Taylor’s theory. However, I don’t think that such a trick is necessary. On the contrary, the ambivalence between the analytic approach and the intuitive, between his Platonism and the immanent narration of the self, has to be sustained if one wants to take Taylor’s theory seriously. And the key to sustaining it—maybe even without falling into (too much) incoherence—lies in his understanding of Frege.

In the following two sections, I want to show what the gap that eventually is between the (Platonic) theoretical framework Taylor needs to tell us his story of the self and the actual story itself means and how it might be bridged. To this end, I will first take a closer look at Frege’s logic of predication as well as Taylor’s interpretation of it. Thereafter we will see how it fits his philosophy of the self and what it means for his concept of ‘reality.’ Comparing Taylor’s understanding of Frege with the second heavyweight he emplaces, the picture he draws will become clearer: it is the apparently contradictory philosophy of the late Ludwig Wittgenstein that will show us exactly what Taylor means when he talks about ‘reality,’ and what kind of reality that is. In the second section, we will see that the ambivalence is indeed an (implicit) attempt to overcome the old problem of universals in which the Platonic framework, with its universalism, frames the Wittgensteinean narrative with its nominalism. ^{Footnote 11} In the last section, I will briefly summarize the results of my analysis and try to put them in the wider context of Taylor’s philosophy. There we will see that the constitutive goods have been a Platonic framework for Taylor’s otherwise nominalistic theory of the self all along. They are the guarantee of the reality of the reality. They guarantee that ‘all true beliefs are true,’ not just ‘because things are as they are’ but, rather, because of their Platonic structure.

2. Taylor’s Frege: The Context Principle, Sense, and Reference

To understand Taylor’s realism as he lays it out in The Language Animal, we must first understand Frege’s logic of predication. Even if the main topic of Taylor’s latest book seems to be a comparison of what he calls ‘HLC-theory’ and ‘HHH-theory,’ criticizing the former and arguing for the latter, ^{Footnote 12} it is impossible to ascertain his concept of ‘truth’ without Frege. Neither Johann Georg Hamann nor Johann Gottfried Herder nor Wilhelm von Humboldt had a philosophical concept of ‘truth’ (in a strict sense). In fact, Herder in particular, who has been the most influential of them on Taylor’s thinking, ^{Footnote 13} has to be seen as a critic of ‘truth’ as a universal concept. ^{Footnote 14} But, as Taylor repeatedly insists (most of all in his criticism of Foucault), there has to be ‘a something’ that makes our beliefs true (or false). The alternative would be relativism, which he rejects emphatically. ^{Footnote 15} However, what this ‘something’ is that makes our beliefs true (or false) can’t just be something in our heads, as Taylor imputes to Thomas Hobbes, John Locke, and Étienne Bonnet de Condillac, as well as to all post-Cartesian theorists. Instead, it must be something the reality of which is independent of us, of what we believe it to be. And for that Taylor needs Frege. ^{Footnote 16}

Like Taylor, Frege is an anti-psychologist who—in contrast to the former—established a philosophy (of language) in which every sentence, every predication can be split up into a proposition and its subject. In order to be true, the subject of the predication simply has to fulfil its proposition. Thus, the truth of the sentence ‘This is an article about Taylor’s concept of “reality”’ ^{Footnote 17} is evaluated by the subject ‘this’ fulfilling the proposition ‘is an article about Taylor’s concept of “reality.”’ To put it in a more abstract form: the truth of a predication F(x) is determined by the subject (x) fulfilling the proposition F. This is the most basic principle of Frege’s predicate logic. And it basically also already covers what the context principle says: there is no truth—no meaning after all—outside a predication. Neither is ‘this’ able to be true (or false) nor is any other isolated word. Only in the context of its proposition can a word, which will, then, have been a subject all along, be evaluated as making the predication true (or false). ^{Footnote 18} This is what Taylor refers to when he writes, “A word only has a meaning (reference) in the context of a sentence.” ^{Footnote 19} As a matter of fact, his interpretation of Frege’s context principle exceeds the basic predication of a subject (x) regarding its proposition F. To illustrate how and why this is the case, we must clarify another part of the Fregean logic first: the distinction between ‘sense’ and ‘reference.’

In Taylor’s interpretation, it seems not only as if a word has meaning in the context of a sentence but as if, in the context of a sentence, there is no such thing as a meaningless word. At least, Taylor does not mention this possibility at all and he chooses his examples accordingly. Maybe there is no room for meaningless predications in Taylor’s theory because of its very favour for the meaningfulness of the (common) sense. And perhaps it is just because he cannot imagine a predication that doesn’t mean anything, that doesn’t refer to an actual reality. Hence, he postulates that “we get something through a reference and then we predicate something of it.” ^{Footnote 20} But, of course, not every predication, not every sentence, refers to something. And, as for Frege, it does not need to refer to something to make (at least) sense. ^{Footnote 21} Let me give an example to show where and how Taylor exceeds Frege in this respect.

According to the basic principle of predicate logic, every sentence has the form F(x). And if the (x) fulfils its F the sentence is called true. However, what about the sentence ‘The present king of France is bald’? Put in the simple form F(x), we get a ‘present king of France’ that must fulfil his ‘baldness’ to make the sentence true. Certainly, France has currently no king who is able to be bald. ^{Footnote 22} So what is the meaning of this predication? What does it refer to? The answer is evidently nothing. There is no ‘something’ we got ‘through a reference’ of which we ‘then predicate something.’ All there is, is a correct sentence that makes perfect sense. And it does this just by purporting that there is a ‘present king of France’ who is (or is not) ‘bald.’ For Frege, such a predication is doubtless possible, although because it lacks an existing (x) that might (or might not) fulfil its F, it is neither true nor false. ^{Footnote 23} For Taylor, on the other hand, in the context of its F, every (x) is, and must be, able to make the predication either true or false. For him, every sense has (to have) a meaning, a reference. Or to put it the other way round: if we have (always) already gotten ‘something through a reference,’ it is just consequent that the sense of what we have (always) already gotten also has a meaning.

To my mind, the consequences of this interpretation for the concept of ‘reality’ are quite clear. Focusing on the context principle, exceeding its purpose of preventing from the mistake of asking about the meaning of an isolated word and to that effect mingling the Fregean distinction of ‘sense’ and ‘reference’ with what is called the ‘presupposition’ of a word ^{Footnote 24} Taylor establishes a reality in which the truth-valuating context of a word is far more than just its proposition within a predication. ^{Footnote 25} The context that makes a word true (or false) is the guaranteed presupposition in the context of its proposition, its guaranteed representation of an existing reality of which something can be predicated. So the framework of Taylor’s narration of the self is, in fact, one that uses the Fregean logic of predication. But in addition to establishing a philosophy (of language) that makes it possible to evaluate truth without referring to something in the mind, to some belief about ‘truth,’ he uses it to establish a fundamental truth of the things as they (really) are. In contrast to Rorty’s above-mentioned description of Taylor’s realism, I would, thus, rather describe it as shrewdly imprinting meaning on the (common) sense by presupposing the existence of every-thing that is a subject to a proposition, of every (x) of which something is predicated of.

3. Taylor’s Wittgenstein: Presupposition of Meaning and the Language Game

Before we return to the constitutive goods, their Platonic origin, and their actual role as guarantor of the meaningfulness of the (common) sense, we must take a look at Taylor’s interpretation of Wittgenstein. As a pioneer of the Ordinary Language Philosophy, this second heavyweight initially seems quite contradictory to Frege’s predicate logic. But in the light of Taylor’s focus on the context principle—especially in his somewhat peculiar interpretation of it—the Wittgensteinean ‘Language Game’ might paradoxically be exactly what we need to somehow understand why and how every (common) sense has (to have) a meaning and, as a result, to understand Taylor’s concept of ‘reality.’ As much as Taylor insists on there being ‘a something’ that makes our beliefs true (or false), he insists on the narrative dimension of the self, on it being a quasi-itself-telling-story. ^{Footnote 26} Besides the apparent consequence of the self, thus, being itself absolutely true, something that can’t be anything but an (in itself) coherent and completely authentic substance, this combination inevitably leads to an understanding of the Wittgensteinean Language Game as having a ‘true meaning’ in a Fregean manner. Thus, it shouldn’t surprise us that Taylor dedicates one of the concluding chapters of The Language Animal to the argument for “How Narrative Makes Meaning.” ^{Footnote 27} It is the significance of life itself as a narrative process that he wants to found by binding the ordinary, ‘the trivial’ to use Rorty’s characterization, by binding the (common) sense to a reality that ultimately leaves no doubt about its very meaningfulness, its reality.

In a way that is similar to the context principle, according to the Wittgensteinean Language Game—as Taylor quotes it—“words only have the meaning they have within the ‘language games’ we play with them.” ^{Footnote 28} At first glance, this might be mistaken as the exact fundament on which we stand with Frege. On second sight though, the tremendous differences are obvious. While, for Frege, the context in which a word has meaning is its proposition and what is true (or false) about it is its fulfilling (or not fulfilling) the proposition, for Wittgenstein the context is its conventional use in a language. ^{Footnote 29} Therefore, one cannot even talk about ‘meaning,’ not to mention ‘truth,’ in a Fregean manner. Instead, words only make sense in the way they are spoken and understood. They don’t actually refer to something beyond their role within the language, beyond their purpose within the game in which they are used. Nevertheless, for Taylor, there are apparently still enough parallels to compare, and indeed combine, both theories.

One of the most impressive examples that illustrates how the Wittgensteinean Language Game works is, in my opinion, that of an apprentice being sent to fetch breakfast. As he returns with sandwiches and coffee, his fellows unexpectedly get angry with him and ask why he has not brought breakfast but has, instead, wasted all the money on food and coffee. ^{Footnote 30} The trick of the matter is them being craftspeople for whom demanding breakfast ‘means’ demanding beer. The apprentice simply ‘referred’ to another language game with other rules than those used by his fellows because he did not know the ‘right’ language game with the ‘right’ rules and therefore got the ‘meaning’ of ‘breakfast’ wrong. Compared to Frege, there are a lot of things that work differently here. (1) There is obviously no such thing as a universal truth the word ‘breakfast’ could possibly fulfil by being subjected to a proposition. (2) The ‘meaning’ of the word ‘breakfast’ isn’t something we get by referring to ‘a something’ but through its use and the success of the action that is associated with it. (3) The context that provides ‘meaning’ to the word ‘breakfast’ is not one of trueness but rather one of adequacy. (4) The only difference between the sense and the reference of the word ‘breakfast’ is the former being its understanding and the latter it being understood, i.e., its proper usage.

However, what if we, like Taylor, implement this Wittgensteinean view of language within the greater framework of the Fregean predicate logic (respectively framing the former by the latter)? What would this do to the context that defines whether a word is used correctly? What would it do to the sense of a word which is its understanding, to the meaning that is its being understood, being properly used? By expanding the Fregean truth-valuating context F “beyond this [the actual proposition] to consider what larger texts [the narrative as such] can show us about language and its powers,” ^{Footnote 31} Taylor creates a concept of ‘reality’ in which the given narrative, the language game with its rules and conventions and with its correct usage, functions as one great truth-defining predication in a Fregean manner. That is how finally the ordinary, the trivial, the (common) sensical ‘narrative makes meaning.’ While with Frege alone, there is no truth—no meaning after all—outside a predication, with Taylor’s interpretation of the context principle and his inflationist approach to truth (which he accomplishes by implementing Wittgenstein in Frege), the correct usage of a word alone is, as strange and tautological as it may sound, what provides its true meaning, what evaluates its trueness within its predication, which is the whole narrative as such.

Admittedly, this sounds a lot like Rorty’s sarcastic description of Taylor’s realism quoted above. Nonetheless, I believe there are a few new insights here. As a matter of fact, the triviality, the commonness of Taylor’s realism, appears not to be that uninteresting after all. If anything, it is exactly what Taylor intended: the foundation of the significance of life itself as a narrative process by binding the ordinary, the trivial, the (common) sense, to a reality that ultimately leaves no doubt about its very meaningfulness, its reality. And he does this by imprinting the Fregean ‘meaning’ on the Wittgensteinean (common) sense. What is still missing in order to really ascertain Taylor’s concept of ‘truth’ and ‘reality’ is ‘a something’ that guarantees the actual reality of this ‘Fregensteinean’ cosmos as such. By achieving ‘a something’ like that, something that presupposes every subject to that ‘holistic predication’ as an always already existing subject of that very ‘holistic predication’—which would be nothing less than the self—the mentioned gap might be closed in a manner that (if we bite the bullet) also overcomes the old problem of universals.

4. Taylor’s Constitutive Goods: The Reality of the Reality and the Necessity of being Oneself

In conclusion, Taylor’s realism does indeed try to unify a radically immanent narration of the self, its becoming through its ‘cultural reality’ and its not being framed by anything other than its being-in-the-world, with a concept of ‘truth’ that is clearly inflationary. Besides the former commonly applying only to relativist theories that appear to contradict the latter, the most interesting aspect of Taylor’s attempt might be the fact that neither Frege nor Wittgenstein can be labelled as inflationists in their own right. It is only in their combination that the Wittgensteinean narration receives its ability to be ‘true’ at the same time as the Fregean predication receives its ability to expand to that holistic one that then eventually functions as some sort of ‘logic of narration.’ Such an expansion of Frege’s otherwise cautious, almost reluctant, theory as far as the ‘ordinary’ is concerned certainly needs a great amount of assurance. And this assurance is provided by the constitutive goods as Taylor uses them leastwise since his Sources of the Self. By being not just ‘the content of the moral theory’ (which is the Wittgensteinean narrative) but ‘a something the love of which empowers us to do and be good’ (which is the structuring of the content after the Fregean predicate logic), the constitutive goods inevitably inflate the truth—valuating context from a proposition within a predication to the rules or the conventions (the Wittgensteinean Language Game as such) of the given narrative.

Because of the constitutive goods, the truth-defining context of every word is always already its convention, its usage according to the rules. That is the reality of the reality; by the Platonic guarantor, the constitutive goods guarantee reality of everything that is a subject not only to, but of its ‘holistic predication,’ its narrative. To put the point plainly, everything true is true because it is structured by, after and according to the reality of the narrative that is guaranteed by the constitutive goods.