The philosopher Stuart Hampshire died on June 13th. According to this obituary he was highly regarded as a moral philosopher. I'm not qualified to comment, but he was clearly an important member of the loose group of philosophers at Oxford including Austin and, particularly, Grice, that a lot of pragmatics has its roots in. (Some other names are mentioned in this post in the Relevance list archives.)
In Grice's Aspects of Reason Hampshire makes a cameo appearance as "the hapless Shropshire in Chapter 1 who "reasons" in one quick step from the premiss that chickens run around after their heads are cut off to the conclusion that the human soul is immortal." as Richard Warner puts it in his introduction (available as a pdf here).
Grice was making a serious point: it's not good enough - if we aim to define reasoning (or good reasoning) to count all cases like Shropshire's "argument" where the conclusion follows from the stated premise plus a number of unstated premises. On the other hand, we want to count some such cases - there are plenty of good arguments in which the conclusion only follows from the stated (or mentally entertained) premise or premises when they are combined with unstated (or non-entertained) premises. Grice certainly didn't want to have a definition of reasoning which excluded these cases.
One important consequence of this for Grice's pragmatics is the way it bears on the debate about whether conversational implicatures are always calculated by explicit reasoning or, at least sometimes, via heuristics or shortcuts.
The discussion around the Shropshire example in Aspects of Reason suggests that Grice could choose both options simultaneously: the derivation of a conversational implicature need not be thought through on any particular occasion to count as having been produced by (sound) reasoning, since in general, sound reasoning does not involve consciously entertaining all of the steps in a logical derivation.
Therefore (and not coincidentally) the criterion in Aspects... for sound reasoning ends up being very similar to the idea that conversational implicatures have to be derivable from what is said plus other premises, so a rational reconstruction should always succeed. What the discussion of Shropshire adds to this is that not just any proposition is to be allowed as an unarticulated premise in this kind of reconstruction. It is one of the central problems of pragmatics to explain which extra premises are allowed in which cases.