The Invisible Constitution in Comparative Perspective

language is part of an activity, or a form of life . . .

‘So you are saying that human agreement decides what is true and what is 

false?’ – It is what human beings say that is true and false; and they agree in 

the language they use. That is not agreement in opinions but in form of life . . .  

What has to be accepted, the given, is – so one could say – forms of life.

28

Wittgenstein goes so far as to say that ‘If a lion could talk, we could not under-

stand him’.

29

 Although somewhat aphoristic, Wittgenstein is presenting the 

core ideas of the conventionalist picture, but extending it beyond formal axi-

omatic systems to practices of language use and the following of rules more 

generally.

30

27 

A. J. Ayer, Language Truth and Logic 2nd edn (London: Victor Gollancz, 1946) 105–6.

28 

Ludwig Wittgenstein, Philosophical Investigations (Oxford: Blackwell Publishing, 1953), §§ 19, 

23, 241, 226 (italics in original).

29 

Ibid.

, 223.

30 

For a sympathetic discussion of Wittgenstein’s approach to rules of inference (referring to 

his  Remarks on the Foundations of Mathematics rather than the quoted passages from the 

 

---

158 

Patrick Emerton

When it comes to logic and mathematics, it is generally accepted that the 

language (English, French, Japanese, etc.) and broader cultural context in 

which an axiomatic system is set out, and in which consequences are derived, 

does not affect the validity of the reasoning. To borrow Wittgenstein’s termi-

nology, whatever exactly it is about human ‘forms of life’ that enables us to 

establish meaningful logical terms and thereby rules of inference that under-

pin valid arguments, these forms of life seem to be common across human 

communities.

31

 But logic and mathematics are in some ways very simple and 

primitive elements of human experience, at least in their foundations (nega-

tion, if  . . .  then  . . . reasoning, counting, measurement, etc.), and hence it may 

not be that surprising that they are common across human beings. When 

it comes to the key terms that figure in a rule of recognition, however, and 

hence to the rules of legal inference that (on this picture) would be implicit 

in them, we might expect a greater degree of variation in the relevant forms of 

life. Thus, in addition to necessarily being unexpressed, there is good reason 

to think that on the ‘thick’ picture of validation the rules of legal inference 

must be, in some sense, ‘local’ or ‘parochial’ rather than universal. That is 

not to say that different constitutional orders must be thought of as mutually 

incomprehensible Wittgensteinian lions, but on the ‘thick’ picture we might 

expect constitutional orders to differ from one another in their legal inferential 

practices in ways that do not emerge self-evidently from their written constitu-

tional texts, and for reasons that would not figure in a typical catalogue of the 

elements of that legal order.

32

 Hence we have good reason to think that, on the 

‘thick’ Hartian conception of validation, every constitutional order necessarily 

includes an important invisible aspect.

Philosophical Investigation), which also relates this approach to the resolution of Lewis Car-

roll’s puzzle, see A. B. Levison, ‘Wittgenstein and Logical Laws’ (1950) 14 Philosophical Quar-

terly 345, 348–51.

31 

For one account of these linguistic capacities in terms of common human cognitive capacities, 

see Stephen J. Barker, Renewing Meaning: A Speech-Act Theoretic Approach (Oxford: Claren-

don Press, 2004), chapters 7 and 8.

32 

Brian Bix has rightly cautioned against overzealous use of Wittgensteinian argumentation 

to address questions that arise in the legal domain, which is quite unlike the mathematical 

and logical domains with which Wittgenstein was chiefly concerned: ‘The Application (and 

Mis-Application) of Wittgenstein’s Rule-Following Considerations to Legal Theory’ in Law,