This is the review I wrote on Amazon.com:

Even though Wittgenstein’s German is nothing like Kant’s, providing a good translation of his work is a challenge given all that one must bring into consideration. Anscombe’s original translation had its merits, but it also had a number of frustrating flaws.

One of the many problems with Anscombe’s translation of PI, is her translation of both “hinweisende Erklärung” and “hinweisende Definition” as “ostensive definition,” where the former is more literally read as “ostensive explanation” and the latter as “ostensive definition.” See, e.g., §§27 and 28 of an earlier edition. And as one can see from Wittgenstein’s discussion, there are times when he uses “hinweisende Erklärung” to mean “ostensive explanation” as opposed to actually ostensively defining a word, e.g., §31. And sometimes he uses them together almost interchangeably, e.g., the last two lines of §28. One of the most glaring cases of Anscombe ignoring the distinction is in §6 where the German reads, “Dies will ich nicht `hinweisende Erklärung’, oder `Definition’, nennen….” and the English translation reads simply “I do not want to call this `ostensive definition’….”

One way this difference, and Anscombe’s failure to track it, is important is that giving an explanation is a much more open ended activity than giving a definition in a somewhat similar way as the German word for “game,” “das Spiel,” is more open than the English word, since “das Spiel” can also mean the more open concept of play.

One small “problem” presented by the updated translation is that the changes make past expressions no longer so apt, e.g., talk of a “no stage-setting” interpretation of the failure of the private ostensive definition in §258, based on the remarks about stage-setting in §257, is now problematic, since the new translation does not make use of the expression “stage-setting.” This is a small problem, however.

While I respect Hacker’s work, I do not agree with how easily he attributes substantive views to Wittgenstein; so I worry about how Hacker’s methodological assumptions about Wittgenstein influence his input on the revisions. Nevertheless, I do not have a similar worry about Schulte, and I know that both Hacker and Schulte took into consideration the suggestions of other Wittgenstein scholars when making the revisions.

It is too soon to tell now, but I am excited to see what kind of an effect this new edition has on Wittgenstein studies.

Philosophical Investigations, 4th Edition