Tag Archives: Austin

In a footnote to ‘A Plea for Excuses,’ J. L. Austin wrote the following:

Plato, I suppose, and after him Aristotle, fastened this confusion upon us [the confusion involved in collapsing succumbing to temptation into losing control of ourselves], as bad in its day as the later, grotesque, confusion of moral weakness with weakness of will. I am very partial to ice cream, and a bombe is served divided into segments corresponding one to one with the persons at High Table: I am tempted to help myself to two segments and do so, thus succumbing to temptation and even conceivably (but why necessarily?) going against my principles. But do I lose control of myself? Do I raven, do I snatch the morsels from the dish and wolf them down, impervious to the consternation of my colleagues? Not a bit of it. We often succumb to temptation with calm and even with finesse. (Austin 1956: 146 (reprint))

Austin appears partly to be motivated here by the idea that losing control of oneself must involve total, or near total loss: failure to govern oneself in accord with one’s views or intentions concerning what is to be done. If one allows that any failure in self-governance, any slack between one’s views about what is to be done and that which one in fact does, counts as a loss of control, then the proposed distinction between succumbing to temptation and losing control becomes less clear. However, Austin seems right to claim that one can act in accord with one’s then operative views about what is to be done, and to that extent count as fully in control of oneself, whilst counting as having succumbed to temptation, due to the fact that one’s then operative views about what is to be done conflict with what is in fact to be done, or with what one’s principles permit. There is, as ever, more to say here. However, I wish instead to consider a distinct case of bombe disposal.

Consider the following case. The bombe has been distributed fairly so that each diner has had a segment, but one segment remains. You desire the segment, and set about deliberating on the question whether to take it. You have the following views:

(V1) You are permitted to take the final segment of ice cream only if you are entitled to the final segment of ice cream.

(V2) No one who would decide to take the final segment of ice cream is entitled to the final segment of ice cream.

According to (V1) and (V2), it would go against one’s principles to take the final segment of ice cream on the basis of having decided to take the final segment of ice cream. By (V2), anyone who decided to take the ice cream would thereby lose their entitlement to take the ice cream; and by (V1), one is permitted to take the segment only if one is entitled. However, in advance of your deciding to take the final segment, (V1) and (V2) seem to leave open that you may take the ice cream. And they thereby seem to leave open that you can decide to take the ice cream. So, can you decide to take the final segment of ice cream? More precisely, can you decide rationally to take the final segment without giving up, or losing track of, your standing commitment to (V1) and (V2)?


J. L. Austin (1956) ‘A Plea for Excuses.’ Proceedings of the Aristotelian Society, supp. vol. Reprinted in his Philosophical Papers. Oxford: Clarendon Press, 1975.


Despite the patronage of Bertrand Russell and Ludwig Wittgenstein, Gottlob Frege’s work had surprisingly little direct impact on British philosophy until around the early 1950s. Even amongst British philosophers who actively engaged with German philosophy from around the same period—for one important example, Gilbert Ryle—Frege’s work seems to have figured mainly indirectly before this time. That changed in the 1950s as translations of Frege’s work began to become widely available, and philosophers like Elizabeth Anscombe, Michael Dummett, and Peter Geach began to engage seriously with it. Two key moments were J. L. Austin’s translation of Die Grundlagen der Arithmetik (1884, translated as The Foundations of Arithmetic), published in 1950, and Peter Geach and Max Black’s Translations from the Philosophical Writings of Gottlob Frege, published in 1952, and including, amongst other important pieces, translations of ‘Über Begriff und Gegenstand’ (‘On Concept and Object’) and ‘Über Sinn und Bedeutung’ (‘On Sense and Reference’). This note concerns the former, Austin’s translation of Frege’s Grundlagen.

According to Geoffrey Warnock (1973), the Grundlagen was one of the texts read by Austin’s ‘Saturday Mornings’ discussion group, and, although Warnock suggests that the book was read in translation after 1950, the opportunity to discuss Frege’s work in the group may have played a role in Austin’s decision to produce the translation. In an encyclopedia entry on Austin’s work, I suggested that Austin’s translation was produced ‘so that it could be set as an exam,’ but I failed to record a source for the remark. The source was Michael Dummett, who records the connection in the following passage:

My fascination with the writings of Frege dates from my reading, as an undergraduate, of the Grundlagen der Arithmetik, unquestionably the most brilliant sustained performance of its length in the entire history of philosophy; and, as I then knew no German, this was made possible by Austin’s translation of that book, which first introduced it to most English-speaking philosophers at a time when there was very little interest in Frege, and was occasioned by its inclusion, I believe at Austin’s suggestion, as one of the of the texts to be studied for an excellent optional paper in the Oxford Philosophy, Politics and Economics Honours School. (Dummett 1978: xxiii–xxiv)

Dummett is clear here that the production of the translation was occasioned by the inclusion of the Grundlagen as an examined text, although he leaves open whether its inclusion as an examined text was also due to Austin and also whether the availability of a translation was a necessary condition of its inclusion. However, Dummett’s claim, together with his admitted inability at that time to read German, makes plausible that the availability of a translation was a necessary condition of reasonable inclusion. Dummett says a bit more in the following two passages:

It just so happened that Austin did a very good thing by inventing an optional paper in P.P.E., which I read, which was called absurdly, ‘Foundations of Modern Epistemology’, and consisted of a number of set texts, starting with the Theaetetus, and finishing with Frege’s Grundlagen. It was for that purpose that Austin translated the Grundlagen. (Dummett 1993: 169)

I’ve always remained an analytic philosopher—but as for logic and philosophy of mathematics, that’s a separate thing. It happened, well again, quite accidentally. I took, the first time it was set, an optional paper in philosophy in my final examination. It was one invented by John Austin and it was called, absurdly, The Origins of Modern Epistemology. What it was was a collection (a rather large collection) of texts, starting with Plato’s Theaetetus and finishing with Frege’s Foundations of Arithmetic. These were texts one wouldn’t normally have come across during the ordinary Philosophy, Politics and Economics course in Oxford, and I worked my way through these. I was very interested in a lot of them but I was absolutely bowled over by the Foundations of Arithmetic, and I thought, I want to read everything this man has written. (Fara and Salles 2006: 2)

Daniel Isaacson adds some further information in the following passage from his obituary for Dummett:

In Finals, in Trinity Term 1950, he took a paper “invented by John Austin” for first examination in that term called “The origins of Modern Epistemology”. Candidates were expected to study four texts from a list of seven, one of which was Frege’s Grundlagen der Arithmetik, newly translated by Austin for this purpose. The examiners’ report records that seven candidates took this paper and that Boole and Frege “attracted the least attention”. One can infer that perhaps only one or two candidates studied Frege for this exam. Nonetheless, there was a class on Frege’s Grundlagen in Hilary Term 1950 that met twice a week, given by Mr. W. Kneale and Mr F. Waismann. Dummett’s ensuing work on Frege has transformed understanding of Frege’s philosophy. Dummett wrote recently of Frege’s Grundlagen der Arithmetik, “I thought, and still think, that it was the most brilliant piece of philosophical writing of its length ever penned.” (Isaacson, ‘In Memoriam: Michael Dummett (1925-2011)’)

This post was occasioned by a question from Michael Kremer. I’m grateful to Michael, and also to Michael Bench-Capon and Aidan McGlynn, for help in assembling sources. Thanks also to Robert May for comments and questions that led to corrections.


Michael Dummett (1978) Truth and Other Enigmas. London: Duckworth.

Michael Dummett (1993) Origins of Analytical Philosophy. London: Duckworth.

Rudolf Fara and Maurice Salles (2006) ‘An Interview with Michael Dummett: From Analytical Philosophy to Voting Analysis and Beyond.’ [Online]. London: LSE Research Online.

Geoffrey Warnock (1973) ‘Saturday Mornings.’ In Isaiah Berlin ed. Essays on J. L. Austin. Oxford: Clarendon Press.

1. According to J. L. Austin, “It takes two to make a truth.” (1950: 124 fn.1.) More emphatically, “When a statement is true, there is, of course, a state of affairs which makes it true.” (1950: 123.) Julian Dodd quotes this fragment and comments:

“Here we see Austin endorsing what has come to be known as ‘the truthmaker principle’, which we may formulate (provisionally, at least) as saying that whenever something is true, there must be some thing whose existence guarantees its truth. It is, however, the ‘of course’ that is particularly revealing, since it sees Austin putting his finger on what one might justifiably regard as a prevailing mood in analytical philosophy. For Austin is claiming that the truthmaker principle is not just true, but obviously true. It is something which no self-respecting philosopher would wish to deny.” (Dodd, 2002: 69.)

Predictably, Dodd does deny the principle. More carefully, he thinks that the principle he claims to find in Austin has not yet been justified, and doubts that it can be.

I think that Dodd is right to connect Austin’s discussion with more recent discussions of issues surrounding the truthmaker principle. However, I’d like to raise some questions about the precise way in which Austin figures in Dodd’s narrative.

(I don’t wish to enter here into the details of Dodd’s negative discussion of the truthmaker principle. I note, however, that, whether or not Dodd is right to reject the various considerations in favour of the principle that he considers, it is hard to see how they could have figured as motivations for Austin’s claim. For, as Dodd emphasises, Austin appears to have taken his claim to be obviously correct. And none of the considerations that Dodd considers has the patina of manifest acceptability.)

The three questions that I shall consider are these.

Q1. Was Austin really committed, in the remarks that Dodd quotes, to a version of the truthmaker principle?

Q2. If Austin was so committed, what might have motivated him to be so?

Q3. Is the proposed motivation something that someone might reasonably take to be obvious?

2. Let’s begin with Q1. There are at least two reasons for thinking that a commitment to the truthmaker principle cannot be read-off from the remarks of Austin’s that Dodd quotes. The first reason is that, taken in context, Austin’s focus isn’t on statements per se—or, more generally, on truth-bearers per se—but rather on statements (truth-bearers) engaged in response to what is observable. Here, there are three central pieces of evidence:

(i) Austin’s most general characterisation of the account of truth that he wants to defend is this:

“When is a statement true? The temptation is to answer (at least if we confine ourselves to ‘straightforward’ statements): ‘When it corresponds to the facts’. And as a piece of standard English this can hardly be wrong. Indeed, I must confess I do not really think it is wrong at all: the theory of truth is a series of truisms. Still, it can at least be misleading.” (1950: 121)

Reformulating slightly, we have that “when (if) a statement corresponds to the facts, (then) it is true.” Clearly, that provides only a sufficient condition on statement-truth and not, in addition, the necessary condition that Dodd finds in the later remarks from which he quotes.

(ii) Austin’s immediate development of his general characterisation makes appeal to what he calls demonstrative conventions. (1950: 121–2.) Austin’s idea is that, as part of one’s making a statement apt for assessment as true or false, one must select an area of one’s environment, or elements of one’s environment, as the area, or elements, who’s antics are to bear on whether one’s statement is true or false. It’s natural to think that such selection will typically be based on immediate perceptual contact with areas or elements, as is suggested by the characterisation in terms of what one is in a position to demonstrate to an audience.

(iii) That leaves open what to say about statements whose truth seems not to depend on an area or elements that it makes sense to think might be demonstrated. For example, it leaves open what to say about statements—or what are apparently statements—that generalise more widely than immediate perception can reach—for example, a statement made using “all swans are white”—or have non-perceptible subject matters—for example, a statement in pure arithmetic. With respect to such putative statements, Austin wavers between treating them as apt only for assessment along dimensions other than the true–false dimension, and treating them as true in a way that doesn’t perfectly fit his basic account. Although the former option would be consistent with Austin’s taking his basic account of statement-truth to apply across the board, the fact that he doesn’t see the second option as immediately foreclosed strongly suggests that his main explanatory target is the truth of statements about what is perceptible.

3. That’s the first reason for taking the evidence to which Dodd appeals to be more equivocal than Dodd suggests. The second reason for doubt is that the transition from what Austin says to the principle, or principles, that Dodd attributes to him is non-trivial. The transition in question is from (A), which Austin explicitly commits to, to (D), which Dodd attributes to Austin:

(A) If a statement is true, then there is a state of affairs which makes it true.

(D) If a statement is true, then there is a thing which makes it true.

The most obvious way of making the transition would be via appeal to a principle like the following:

(S=T) States of affairs are things.

As far as I can see, Austin doesn’t explicitly commit to such a principle, so it remains an open question whether he would have endorsed the transition that Dodd attributes to him. Here, we must beware in particular a slide between three readings of “things”

(Thing 1) Things are the values that must be assigned to quantifiers in sentences used to make true statements in order to characterise the conditions on which true statements can be made by the use of those sentences.

(Thing 2) Things are the values that must be assigned to first-order quantifiers in optimal regimentations of sentences used to make true statements in order to characterise the conditions on which true statements can be made by the use of those regimentations.

(Thing 3) Things are all of the same metaphysical kind, a kind whose most paradigmatic instances are tables, chairs, cats, mats, and their ilk.

On one natural construal, Austin’s talk of “a state of affairs” involves a form of natural language existential quantification over states of affairs. If that construal is defensible, and if Austin would have endorsed (A) when so construed, then we would have evidence of Austin’s commitment to (S=T) read in accord with (Thing 1). However, in advance of further work, we do not yet have the evidence to detach the consequent of that conditional. Moreover, it is far from clear that all natural language quantification is optimally regimented in first-order form. For example, it is a currently open question whether natural language allows quantification into predicate position (roughly, quantification over things people are or do) or sentence position (roughly, quantification over things that are, or are not, so). So, it is far from clear that Austin meant to commit to (S=T) when read in accord with (Thing 2). Furthermore, it is plausibly false that all natural language quantification is over things that are substantively, or interestingly, like tables, chairs, cats, mats, &c. There is no reason at all to hold Austin to a commitment to (S=T) when read in accord with (Thing 3).

The truthmaker principle can itself be read in accord with any of (Thing 1)–(Thing 3). It follows that, as long as Austin can be convicted of endorsing (S=T) on at least its weakest, (Thing 1) reading, then he is committed to a version of the truthmaker principle. However, if that were so, it would then be important to establish that endorsement of the weak version of the truthmaker principle is problematic, or unwarranted.

There are at least two reasons, then, for questioning Dodd’s attribution to Austin of the truthmaker principle. Furthermore, that attribution comes in a variety of strengths, and it remains an open question whether any version of the principle to which Austin commits is really a target for Dodd’s critique.

4. Rather than pursue those issues further, let’s turn to Q2, understood now as the question, what motivated Austin to endorse (A)? I think the answer to that question is simpler than any of the potential answers that Dodd considers. To a good first approximation, Austin takes it that sentence nominals in natural language denote states of affairs. Thus, the sentences in (1) and (2) are correlated with the sentence nominals in (3)–(6), and the latter sentence nominals can naturally be unfurled into (7)–(10):

(1) The cat is on the mat.

(2) The chair is white.

(3) The cat’s being on the mat…

(4) The chair’s being white…

(5) That the cat is on the mat…

(6) That the chair is white…

(7) The state of affairs of the cat’s being on the mat…

(8) The state of affairs of the chair’s being white…

(9) That the cat is on the mat is a state of affairs…

(10) That the chair is white is a state of affairs…

Moreover, corresponding to the claims that (1) and (2) may be used to make statements that are true, we have the following:

(11) (/The state of affairs of) the cat’s being on the mat is a fact (/is the case/obtains).

(12) (/The state of affairs of) the chair’s being white is a fact (/is the case/obtains).

(13) That the cat is on the mat is a fact (/is the case/?obtains/is an obtaining state of affairs).

(14) That the chair is white is a fact (is the case/?obtains/is an obtaining state of affairs).

Such transformations of sentences like (1) and (2) into their nominal correlates in (3)–(10) appear to be more or less productive: they appear to be sustained by competence with a rule, rather than by special knowledge about individual cases. And although there is something artificial—or, at least, prolix—about some of the alternatives listed in (11)–(14), they present nothing that would give pause to an ordinary speaker of English. And the transitions from (1) and (2) to (11)–(14) seem equally unproblematic. That’s not to say that further philosophical work couldn’t find reason to question any of the listed transformations and transitions. Rather, it’s to suggest that it would be surprising if such reasons were found. And it would plausibly indicate a need to consider modulating what would otherwise be a perfectly ordinary willingness to allow the transformations and transitions.

5. If that’s right, then our ordinary modes of talk about states of affairs align with Austin’s (A). Moreover, it seems that the motivation for (A) sketched here, drawing as it does on ordinary competence with relevant bits of English, supports an affirmative answer to part of Q3: it would be reasonable to take as obvious—albeit, and as ever, defeasibly so—that if a statement is true, then there is a state of affairs to which it corresponds. It would remain open question whether the state of affairs to which a true statement corresponds is such as to make true the statement. But it would be plausible to hold that the obtaining of the state of affairs would suffice for the truth of a corresponding statement.


Austin, J. L., 1950, “Truth,” Proceedings of the Aristotelian Society, Supplementary Volume 24: 111–128. Reprinted in Austin’s Collected Papers, Oxford: Clarendon Press. References to reprint.

Dodd, J., 2002, “Is Truth Supervenient on Being,” Proceedings of the Aristotelian Society, 102, 1: 69–85.

%d bloggers like this: