I am grateful to James Levine and two anonymous referees for their helpful comments on previous drafts of this essay.
1In this essay, I argue that Wittgenstein attempted to clarify ethics through a procedure that, by analogy with “transcendental arguments”, I call “transcendental thought experiment”. The paper is divided into three sections. First (§1), I give a brief perspectival characterization of transcendental arguments inspired by Mark Sacks’ work, suggesting that they are argumentative procedures that require us to entertain a representation in the first-person perspective, in order to recognize necessary conditions of the possibility of representation of the world. Then (§2), I introduce the notion of transcendental thought experiments, arguing that they are imaginative perspectival procedures aimed at recognizing necessary conditions of the possibility of representation of the world, which, as such, need not necessarily take the form of an argument. Finally (§3), I argue that in his Lecture on Ethics Wittgenstein employed a transcendental thought experiment to clarify ethics and (what we may cautiously call) ethical expressions. The result will be a clearer view of Wittgenstein’s own ethical outlook, in the light of his claim in the Tractatus that “Ethics is transcendental” (T, 6.421).
- 1 See Smith and Sullivan (2011, p. 7) and Gardner (2015, p. 2).
- 2 Kant defines cognition as a “the determinate relation of given representations to an object” (B137 (...)
2It is agreed on virtually all hands that Transcendental Philosophy is concerned with necessary conditions of the possibility of cognition.1 Whatever else cognition is, it surely must involve, if not be equivalent with, representation of the world (which latter may then take the name “experience”, “thought”, language”, etc.).2 If so, Transcendental Philosophy must be concerned with necessary conditions of the possibility of representation of the world. Yet exactly insofar as this could be said to be the fundamental concern of Transcendental Philosophy, the problem immediately arises as to how one is to recognize such necessary conditions in the first place.
3According to one rather influential view, in order to recognize and establish the validity of necessary conditions of representation of the world, we need specific argumentative procedures, known as transcendental arguments. The idea of transcendental argumentation goes back at least to Kant (who preferred to talk of “transcendental proof”3), and has fascinated philosophers on and off ever since, becoming the subject of a heated controversy in the second half of the 20th Century, in the light of the work of Peter Strawson, who disentangled transcendental argumentation from transcendental idealism (e.g. 1959 and 1966).
4Nowadays, it is customary to say that transcendental arguments are deductive argumentative devices that have anti-skeptical import, or that anyways aim at establishing the legitimacy of a cognitive claim, by combining the rigour of well-known rules of inference – notably, the modus ponens – with an appeal to the idea of “necessary conditions of possibility” (see Gava, 2019). In other words, transcendental arguments are often taken to be arguments of the following form:
P
If p [the conditioned], then q [its necessary condition]
Therefore, q4
where p is a premise concerning (subjective or objective) representation and/or the cognitive capacities involved in it (e.g. the capacity for experience, thought or language). However, if understood in certain specific ways, this schema of transcendental arguments may easily come under fire.
5One recent critique concerns the talk of “necessary conditions”. In the analytic tradition, this has often been taken as talk of merely analytic necessary conditions – that is, conditions that are to be reached just by logical analysis of propositions (cf. Franks, 1999, pp. 117-8), and whose necessity depends upon the logical structure of our concepts alone (cf. Gava, 2019, p. 454). However, this makes it rather hard to see what is so special about transcendental arguments. Consider for example the following modus ponens:
(I represent that) Tim is a whale.
If Tim is (represented as) a whale, then it is also (represented as) a mammal.
Therefore, (I represent that) Tim is a mammal.
- 5 By this, of course, I do not mean that one needs to have a picture of a mammal (mental or otherwis (...)
- 6 My example is intentionally provocative, but it shows, I think, that it cannot be right to reduce (...)
- 7 It may be objected that the appeal to “possible experience” is an appeal to Kant’s transcendental (...)
6One could surely say that being a mammal is a necessary condition of Tim’s “whaleness”, or for that matter that representing Tim as a mammal is a necessary condition for representing Tim as a whale,5 and that we can know such things out of logical analysis alone. But I doubt most people would want to call this, or at any rate anything even remotely like this, a transcendental argument.6 Not by chance, if Kant was categorical on one thing about transcendental argumentation, it is that transcendental proofs can never be understood in merely conceptual terms, for they essentially contain an appeal to possible experience (A782-3/B810-1).7
- 8 For a critique of the tendency to understand the word “argument” in the phrase “transcendental arg (...)
- 9 The perspectival or first-personal character of transcendental arguments had earlier been acknowle (...)
7Among the Kantian critics of a merely conceptual construal of transcendental argumentation, Mark Sacks (2005) stands out. According to Sacks, as long as we take transcendental arguments as merely deductive procedures, what is distinctly “transcendental” about them gets lost.8 To be clear, Sacks is not against the view that transcendental arguments could take the form of a modus ponens. However, he believes that attention to more than the deductive structure of their presentation is required to follow them (see Gardner, 2015, p. 14). Indeed, according to Sacks, only if we recur to a perspectival schematization of the thoughts or propositions involved in transcendental arguments, then we can see or experience for ourselves that something counts as a necessary condition of the possibility of representation.9
- 10 Indeed, this is the conclusion that Kant aims at establishing in his First Analogy, though not abo (...)
8To appreciate the strength of Sacks’ point, take for example Kant’s “argument for substance” in his First Analogy of Experience, which is one of Sacks’ own examples of a transcendental argument (see Sacks, 2005; 2006). According to Sacks, there is no analytic route leading us from the concept of a ‘change’ employed there by Kant to the concept of a ‘substance’ – the concept, that is, of an abiding something underlying change as a necessary condition of its representation. And indeed, one may analyse her notion of ‘change’ all she likes, but she surely won’t find in there the idea that there must be substance (let alone establish that there must be substance in reality10). If, however, we imagine ourselves as experiencing a certain change, say the change occurring when a lighting hits a tree and reduces it to ashes and smoke, then according to Sacks we can come to see from our own perspective that there must be a persisting substance underlying that change, in order for that entire imaginative procedure to be carried through in the first place. For how could we even imagine experiencing the change at stake, without thereby imagining a thing that was changing all along?
- 11 Differently from Sacks, I do not believe that perspectival thoughts need to be strongly opposed to (...)
9Sacks calls similar perspectival imaginings “situated thoughts” and he claims that they are needed to work our way through the key step of a transcendental argument – the one leading us from the conditioned (e.g. the representation of change) to its necessary condition (e.g. there being a persistent substance). However, the label “situated thought” almost makes it seem as if these thoughts can only be entertained from some specific spatiotemporal location. I will talk more generally of perspectival thoughts – that is, thoughts that are entertained as a result of a self-conscious activation of our cognitive capacities, or in the first-person perspective.11
10In any case, similarly to Sacks, I too suspect that perspectival thoughts are needed to work our way through the key step of at least some transcendental arguments.
- 12 As far as I can make out, there is no body of literature on the general notion I am about to intro (...)
11Here, however, I shall not argue at length for this view. Indeed, for the purposes of this essay, I am not interested in the specifics of Sacks’ proposal, nor for that matter in transcendental argumentation per se. Rather, I have briefly introduced Sacks’ take on transcendental argumentation, and tweaked it a bit, since a perspectival interpretation of transcendental arguments can act as the guiding thread to the crucial notion of this essay, namely that of transcendental thought experiment.12
12So far, we have concerned ourselves with (Sacksian) transcendental arguments. Specifically, we have seen how, in order to bring into view necessary conditions of world-directed representation, (Sacksian) transcendental arguments combine the rigour of well-known rules of inference (e.g. modus ponens) with imaginative perspectival representations (which Sacks calls “situated thoughts”, and which I call “perspectival thoughts”). I now wish to introduce, by analogy with the brief perspectival account of transcendental argumentation given above, the notion of transcendental thought experiment.
13According to Meynell, “[thought experiments] are narratives that are created to prompt their readers to imagine specific fictional worlds, as kinds of situational set-ups that, when you ‘run’, ‘perform’ or simply imagine them, lead to specific results” (Meynell, 2014, pp. 4161-4162, my emphasis). If so, we may define transcendental thought experiments as imaginative devices or procedures aimed at making us visualize a scenario, which, if properly worked out, is to lead us to realize that something is a necessary condition of the possibility of representation of the world, or at any rate that it is involved with such necessary conditions. Through this definition one may easily see that, like (Sacksian) transcendental arguments, transcendental thought experiments have an irreducible perspectival character, insofar as we are to imagine certain scenarios and see – in our imagination – how they play out. Put otherwise, it belongs to the essence of a transcendental thought experiment that it should be capable of prompting a perspectival thought (or: a host of internally related perspectival thoughts).
- 13 On this score, I find it significant that Putnam characterizes his famous argument – against regar (...)
14Thus, transcendental thought experiments remind us of (Sacksian) transcendental arguments in many ways, and indeed, in the light of our discussion above, it would not be hazardous to say that what makes transcendental arguments unique is the fact that, nested within them, there might be transcendental thought experiments.13 There is however cause for keeping the two analytically distinguished, since not any transcendental thought experiment needs to be part of a transcendental argument. That is: as deeply imaginative procedures, transcendental thought experiments might stand on their own, without having to conform to the constraints imposed by this or that specific rule of inference (e.g. modus ponens), or more generally without being reducible to arguments at all.
15This stance is backed, I think, by the (recent) history of the philosophical debate on thought experiments in general. According to one of the dominant views in the relevant literature, championed by Norton, thought experiments are simply arguments in disguise or “picturesque arguments” (Norton, 1996, pp. 334 and 339; cf. also Norton, 2004). That is because, according to Norton, thought experiments could be always “reconstructed” as arguments (ibid.). Recently, however, there have been expressions of growing dissatisfaction with respect to this view. Some such dissatisfaction may be brought back to the following critical observation: even if all thought experiments could be reconstructed as arguments (and that is a big “if”), that would not yet show that they are arguments. (To use an analogy: a pianist may easily “translate” music into musical notation, but that does not surely mean that the musical notation is the music itself).
- 14 Indeed, it is not a coincidence that Descartes’ Meditations are written in the first-person (Soren (...)
16But by far the most important criticism levelled against the view that thought experiments are just arguments is that this view downplays the imaginative (perspectival) character of thought experiments, considering it as merely ornamental – and thereby inessential. In effect, if thought experiments were none other than arguments in disguise, we should expect philosophers to always prefer their regimented and clear “reconstruction” to their “picturesque” perspectival version. And yet philosophers have often aimed at leading us to insights through their thought experiments (cf. Meynell, 2014) – insights that seem hardly detachable from the stark presentation of the thought experiments themselves (cf. Camp, 2009, p. 124), which prompts us to vividly picture a certain scenario (or even: to vividly picture ourselves in a certain scenario). That is the case, for example, with Descartes’ evil demon scenario as well as with Putnam’s brains in a vat one, which are themselves the product of thought experiments.14 To quote Elisabeth Camp:
[B]y describing counterfactual situations in concrete detail, [thought experiments] can trigger a kind of experiential acquaintance that an abstract description misses. (ibid.)
17It goes without saying that an abstract and dry argument will miss this sui generis experiential dimension too – unless it involves an irreducible perspectival element, as is the case with (Sacksian) transcendental arguments.
18The defender of the view of thought experiments as arguments might still object that there are, hidden behind any thought experiment, a host of background assumptions and beliefs, and that those must be suppressed premises, needed to set in motion the machinery of an argument (see Norton, 1996, p. 354). But we need not see things that way. It is true, behind thought experiments there are often background assumptions and beliefs, without which the thought experiment could not be “run” in the proper way. But instead of calling them “premises”, we might simply say that they are the principles that govern the correct generation and understanding of the relevant scenario, imagination of which the thought experiment intends to prompt. Following Walton (1990) and Meynell (2014), I will group such background assumptions and beliefs – as well as tacit rules and conventions, and even psychological tendencies and cognitive capacities – under the label of principles of generation. And I will note that while the principles of generation behind some thought experiments may perhaps be rendered as premises, it does not follow that they must be.
- 15 Cf. Foot (1967), Williams (1973), and Thomson (1971).
19All the observations above will be of the utmost importance for our investigation of Wittgenstein’s own transcendental thought experiment. Before turning to Wittgenstein, however, one last remark is in order. Thought experiments have been employed in a variety of philosophical contexts and with respect to a variety of philosophical problems (and not just, say, with respect to epistemology and the classic epistemological problem of skepticism addressed by Descartes and Putnam). Arguably, a most fascinating use of thought experiments is connected to ethics and ethical problems or quandaries. From Philippa Foot’s “trolley problem” to Bernard Williams’ case of Jim and the Indians, passing by Judith Jarvis Thomson’s case of the unconscious violinist – thought experiments have commanded attention and interest in ethics.15 And while it is debatable what role exactly thought experiments and “imaginary cases” should play in ethics (cf. Dancy, 1985), and what, if anything, they may establish (ibid.), it is nonetheless scarcely doubtable that they may clarify the ethical problems we grapple with. Indeed, we should expect this nice property of thought experiments in ethics to be preserved by transcendental thought experiments in ethics.
- 16 I should stress once more that, in this essay, I consider just one case study in Wittgenstein’s “e (...)
20So far, our discussion has moved at a high level of generality. But we shall now put to the test the very idea of transcendental thought experiment briefly outlined above, by means of a case study. In this section, I will thus argue that Wittgenstein employed a transcendental thought experiment to clarify ethics and (what we may cautiously call) ethical expressions.16
21The thought experiment of Wittgenstein’s that I wish to characterize as “transcendental” was presented during his 1929 Lecture on Ethics, which Wittgenstein delivered in Cambridge, and of which we possess the full text. Almost halfway through the lecture, Wittgenstein claims that no statement of fact can ever be a judgement of absolute value (e.g. an ethical judgement about the moral goodness or evilness of an action). To clarify the claim, he recurred to the following thought experiment:
[S]uppose one of you were an omniscient person and therefore knew all the movements of all the bodies in the world dead or alive and that he also knew all the states of mind of all human beings that ever lived. And suppose this man wrote all he knew in a big book. Then this book would contain the whole description of the world; and what I want to say is, that this book would contain nothing that we would call an ethical judgement. (LoE, p. 45)
- 17 I allude here to remark 5.631 of the Tractatus, in which Wittgenstein imagines writing a book call (...)
22Here, Wittgenstein asks us to imagine the contents of a book, i.e. to entertain a perspectival thought about a book. The author of the book should be imagined as an omniscient human being, pouring into the book all that he knows. Paying homage to a famous remark of Wittgenstein’s, we might call the book The World as an omniscient person would find it.17 This book, Wittgenstein says, would contain the whole description of the world. In other words, the sum-total of its sentences would describe, in the finest detail, all the facts there are, however many facts there are. Still, Wittgenstein adds, the book “would contain nothing that we would call an ethical judgement”.
23At this point, to further clarify what he is after, Wittgenstein asks us to imagine reading the description of a murder in the book (LoE, p. 45). This description would be extremely detailed, depicting all the physical movements of the murderer and his victim as well as all their mental states (ibid.). These, however, would be just facts in the world, and a fact, by itself, is neither good nor evil for Wittgenstein: it is just a fact. Tellingly, Wittgenstein goes as far as saying that “[t]he murder will be on exactly the same level as any other event, for instance the falling of a stone” (ibid.). Put otherwise, in the book The World as an omniscient person would find it, we could find the extremely detailed description of the murder right after that of a falling stone, without ever finding a judgement concerning the moral evilness of the murderous act, or of the person committing the act. For no absolute value – a value that should, in a sense, impose itself by logical necessity – could ever be expressed by propositions according to Wittgenstein, propositions always depicting contingent facts for him (including merely possible facts, which, if they obtained, would be contingent).18
24We come here to the core background “assumption” behind Wittgenstein’s thought experiment, namely his so-called “picture theory” of the proposition, which Wittgenstein presented in his Tractatus. By 1929, the Tractatus was a sensation in Cambridge, and so we should expect that Wittgenstein’s audience was familiar with the “picture theory”, which thereby could lead us to the principle of generation of Wittgenstein’s thought experiment: our own cognitive ability to picture reality by means of language, which the Tractarian remarks usually grouped under the label “picture theory” are exactly meant to elucidate.
- 19 If this were a contingent fact in the world, it should have to be possible for us to picture our o (...)
- 20 Indeed, given Wittgenstein’s famous pronouncements at the end of the Tractatus, the Tractarian rem (...)
25According to the “picture theory”, all propositions work like pictures or models of facts. Facts, however, are for Wittgenstein contingent states in the world – states, that is, which may be other than they are. If so, necessity can never be pictured or expressed by a proposition according to Wittgenstein. Two interesting things follow from this. First, our cognitive ability to picture the world – which is clearly not a contingent fact in the world19 – cannot be expressed by propositions, and thereby the Tractarian remarks that make up the “picture theory” cannot ultimately act as premises of a genuine argument.20 Second, there is no such thing as “ethical propositions” (T, 6.42), i.e. propositions that should depict “necessary”, “absolute” or “non-accidental” values (cf. T, 6.41).
26And here is the crucial point. All we could ever attempt to say about ethics would melt into our hands for Wittgenstein. We would like to say something about “absolute value”, but all we manage to say, if anything at all, is something about contingent matters of fact. Nonetheless, even when we are (made) aware of this, there will be cases in which we will be tempted to use expressions such as “absolute value” or “absolute good” according to Wittgenstein. He himself, in the Lecture, recalls “cases in which I [viz. Wittgenstein] would certainly use these expressions” (LoE, p. 46; my emphasis). One particular case stands out among all others, as Wittgenstein’s “first and foremost example” (ibid.). As he says:
[Whenever] I want to fix my mind on what I mean by absolute or ethical value [...] it always happens that the idea of one particular experience presents itself to me which therefore is, in a sense, my experience par excellence [...] I will describe this experience in order, if possible, to make you recall the same or similar experiences, so that we may have a common ground for our investigation. I believe the best way of describing it is to say that when I have it I wonder at the existence of the world. And I am then inclined to use such phrases as ‘How extraordinary that anything should exist’ or ‘How extraordinary that the world should exist’. (LoE, p. 47)
27Here, Wittgenstein “describes” his own experience of wonder at the existence of the world so that each member of the audience could recall, in the first-person, the same or similar experiences. Put otherwise, once again Wittgenstein is aiming at prompting perspectival thoughts, which could act as a “common ground for [the] investigation”.
28Now, when one reflects upon them, the formulations “How extraordinary that anything should exist” and “How extraordinary that the world should exist”, which Wittgenstein employs at the end of the passage, attempt to put into words the experience of wondering at the existence of the world (that the audience is supposed to recall). They attempt to ascribe to this experience “absolute value”, i.e. a value that is independent of the contingent obtainment of this or that fact, insofar as wondering at the existence of the world is wondering at the world however it may be. It is wondering at its being, not at its being so-and-so.
- 21 In the Tractatus, Wittgenstein famously distinguishes nonsensical pseudo-propositions from sensele (...)
- 22 LoE, p. 49.
29Exactly for this reason, however, the formulations above are nonsensical for Wittgenstein.21 On the one hand, through them we are trying to ascribe “absolute value” to a certain experience or mental state, namely the experience of wondering at the existence of the world. But every experience or mental state, for Wittgenstein, is a fact (something that has “taken place then and there [and] lasted a certain definite time”22). And all facts are contingencies, the complete description of which would be contained in The World as an omniscient person would find it. Thus, we are trying to ascribe “absolute value” to a contingent fact, that which inevitably misses the mark.
- 23 There is a wide literature on Wittgenstein’s remarks on the inconceivability of putative “alternat (...)
- 24 Cf. T, 2.06.
30On the other hand, when we employ formulations like “How extraordinary that the world should exist”, we are also trying to put into words something that we cannot hold to be contingent, namely there being a world. Wittgenstein points out that it makes sense to say we wonder at something that is thinkable as different from how, as a matter of fact, it is (e.g. saying we wonder at the sky’s being blue, rather than white). Yet it does not make sense to say that we wonder at something that, so to speak, has no thinkable alternatives – something the non-being of which we cannot possibly imagine, such as the world23 (as totality of existing and non-existing states of affairs)24. That is because there being a world is a necessary condition of the possibility of saying anything at all, and thereby of saying we wonder at something in the first place. If, per absurdum, there were no world, there would be nothing to talk about, not even experiences of wonder. Indeed, there would not even be those who are supposed to talk, namely ourselves!
31As a result, anything we might want to say about the experience of marveling at the being of the world will inevitably arrive too late. As Wittgenstein writes in his Tractatus:
5.552 The ‘experience’ that we need in order to understand logic is not that something or other is the state of things, but that something is: that, however, is not an experience.
Logic is prior to every experience – that something is so.
It is prior to the question ‘How?’, not prior to the question ‘What?’
32No proposition, not even those that are contained in The World as an omniscient person would find it, could ever express the “What” of the world, i.e. that there is a world. For that is not something we may ever hold to be contingent, but rather the necessary condition upon which even logic depends (WWK, p. 77).
33Here, however, we are interested in the ethical spirit of these considerations, which comes down to a paradox. Faced with the being of the world, we can experience wonder. And this wonder, as we have seen, is a contingent experience or mental state in the world. Yet, insofar as we wonder at the being of the world, and the being of the world cannot be taken to be contingent, then our wonder does not have the features of an experience. Every experience is about this or that contingent fact (e.g. I see the blue sky, I wonder at it, etc.). Talking of the “experience” that something is, the “experience” of something that we cannot hold to be contingent – that is a misuse of language. Paradoxically, we feel that our experience of wonder at the being of the world, to which we want to ascribe absolute or ethical value, cannot be called an “experience” at all. Yet “experience” is the only word we have here. That is why Wittgenstein said that ethics is running up against the limits of language or, as he also said quoting Kierkegaard, running up against paradox (WWK, pp. 68-69).
34Clearly, then, Wittgenstein’s own formulations must be recognized as nonsense: they would not make it into the contents of The World as an omniscient person would find it. And yet, they can reawaken in us the awareness, indeed the wonder, of there being a world, insofar as – by imaginatively reflecting upon them and the urges behind them – we can come to see that there being a world is a necessary condition of the possibility of our saying anything with sense in the first place, and even of our attempting to say something but failing to do so.
35Working our way through Wittgenstein’s thought experiment in ethics, we may thus realize that the being of the world is transcendental, and that each of the facts of the world – each of those facts that could be described in the finest detail in The World as an omniscient person would find it – can be deeply wonderful, not for its being so-and-so, which can be described, but rather insofar as it exemplifies the being of the world, for which we lack words, but which allows us to put into words all that we can put into words. In this sense, while being itself unspeakable for Wittgenstein, ethics pervades the world – all that could be thought or said (T, 6.421). And so, Wittgenstein writes, “Ethics is transcendental” (ibid.) – as is, we may now add, his thought experiment in ethics.
36In this essay, I have argued that the thought experiment that Wittgenstein proposed in his 1929 Lecture on Ethics should be characterized as “transcendental”.
37First, I have discussed the notion of “transcendental argument” and presented a (Sacksian) construal of transcendental argumentation, based on the notion of “perspectival thought”: a thought entertained in the first-person perspective. Then, by analogy with this perspectival construal of transcendental argumentation, I have introduced the notion of “transcendental thought experiment”. I have argued that, notwithstanding some remarkable similarities (e.g. they may both prompt perspectival thoughts), a transcendental thought experiment is irreducible to a transcendental argument, just like thought experiments are irreducible to arguments. To substantiate this view, I have appealed to the notion of “principle of generation”, arguing that the principles of generation of a thought experiment are not necessarily to be cast as suppressed premises, but they may be, for example, cognitive abilities.
38In the last section of the essay, I have presented Wittgenstein’s thought experiment in ethics, in order to characterize it as “transcendental”. I have called attention multiple times to the imaginative perspectival character of Wittgenstein’s thought experiment (not least through the notion of “perspectival thought”). I have argued that its principle of generation is our cognitive ability to picture reality, that Wittgenstein elucidates in his so-called “picture theory” of the proposition. Finally, I have shown how Wittgenstein’s thought experiment in ethics may lead us to realize that there being a world is a necessary condition of the possibility of representation (so that any attempt to put into words the being of the world arrives too late, leading to nonsense in Wittgenstein’s view.) Thereby, I have concluded that Wittgenstein’s thought experiment in ethics is a transcendental thought experiment.
RSS feed
