It is clear – even from such a brief sketch – that this controversy gives rise to huge differences. 1. While Kripke thinks the problem of our lack of foundations should be posed, Baker and Hacker consider this problem a way of exceeding the framework within which we can make sense. And – 2. – while Kripke sees the consequences of a rule as unfixed, but taking shape through community control, Baker and Hacker see them as determined in practice, and incorrigeable from outside. But their agreement is even more striking. 1. Kripke as well as Baker and Hacker deny that the apparant necessity provoked by rules can be justified in terms of a transcendental truth or reality. 2. They also agree in holding on to their confidence in rule following practices by an [p. 121:] appeal to the community, respectively to an undefined concept of we – could it be an implicit form of Rorty's 'we' (1985: 12): "the liberal intellectuals of the secular modern West"? 'We' are not in need of justification. In practice, there is no reason to worry, since rules are given in the practices of our community (or linguistic convention).
   Kripke as well as Baker and Hacker hold that nothing is justified in the end (neither confidence in rules, nor doubt), but that in customary practice, this doesn't lead to problems. Compared to this agreement – which is shared by lots of interpretors of Wittgenstein –, the mutual differences pale into insignificance. The differences even can be explained only in a confused way as long as one tries to hold on to both points of agreement. How can a rule be either unfixed or determined, if the difference between knowing and not knowing the consequences of a rule can't be definitely justified? How can scepticism be either solvable or nonsensical, if the unacceptability of solutions can be very well understood in lots of practices? If one rejects dogmatical criteria for justification and at the same time accepts the priority of a reassuring practice, the controversy is muddled from the start.
   The work of Baker and Hacker does contain a strong and clear refutation of psychological and sociological realism. It is a step forward in not distinguishing any longer between the conceptual reality of a rule (criterion 1) and the linguistic practice of applicating it (criterion 2). But Baker and Hacker are utterly one-sided when they merely extend the apparent self-evidence of practice, pretending that it eliminates sceptical problems. This does not surmount Kripke's paradox, because it doesn't reckon with its practical import. It is true that practical doubt often transgresses the boundaries of sense – but also language does, in practice! (cf. Gadet and Pêcheux 1981).