Grice LOVED G. A. Paul, the author of "Is there a problem about sense data?" and fellow member of Austin's (second) Play Group that met on Saturday meetings (Austin's first play group met on Thursday evenings, but that was before the "phoney" war).
Grice should NOT have a problem about sense data, and the implicature invited by G. A. Paul's question ("Is there a problem about sense data?") is: "No. What makes you think there is. Or worse, who makes you think there is? Witters?"
After all, a sense datum is, simply, whatever is the immediate object of any of the senses, usually, but not always, with the implicature that it is NOT a material object.
Where did G. A. Paul take the expression from?
Although Oxonian to the core, G. A. Paul took it from Royce.
In 1882, Royce, in "Mind" (the journal of the association that sometimes meet jointly with the Aristotelian Society to which G. A. Paul delivered his essay) worte: "What relation does the external reality bear to the sense-datum?".
Royce was intelligently using 'sense datum' in the _singular_. Since surely what is the immediate object of any of the senses need NOT be pluralised in vain.
In 1890, it had become common parlance and James thought 'sense datum' needed no explanation when he uses the expression in his extremely influential "Principles of Psychology" (recall that the "Mind" where Royce introduced the expression was the "Mind" understood as a journal of "psychology and philosophy" (in that order).
"It is no wonder if some authors have gone so far as to think that the sense-data have no spatial worth at all."
This wonder James doubts though, i.e. that a sense-datum has no spatial worth "at all": merely temporal worth.
In the twentieth century, to which G. A. Paul and H. P. Grice belonged, it was in 1912 that B. A. W. Russell in his popular "Problems of Philosophy" writes:
"Let us give the name of ‘sense-data’ to the things that are immediately known in sensation: such things as colours, sounds, smells, hardnesses, roughnesses, and so on."
He is merely providing an partially ("and so on") enumerative definition of the expression first used by Royce back then. Russell also provides a bit of a conceptual analysis: SD is a sense-datum iff it is a thing that is immediately KNOWN in sensation.
By 1938 -- the jazz age -- the term had become fashionable.
Thus W. S. Maugham, while vacationing in the South of France, wrote in Summing Up":
"The sense-datum, on which I thought all KNOWLEDGE was based, seemed to me something GIVEN,"
-- since after all 'datum' is 'given' -- Maugham, who was born in France, can be obvious sometimes --.
Maugham goes on:
"which had to be accepted whether it suited the convenience or not."
To sum up, that is.
In 1956, the once enfant terrible of Oxford philosophy, A. J. Ayer, wrote in Problems of Knowledge (that J. L. Austin will laugh at in his "Sense and Sensibilia", inspiring, if not G. A. Paul, H. P. Grice):
"What is immediately given in perception is an evanescent object called an idea, or an impression, or a presentation, or a sense-datum, which is not only private to a single observer but private to a single sense."
Ayer plays with linguistic botany: sense datum he equates more or less to a Lockeian idea, or a Humeian impression, or a pre-sentation (vs. re-pre-sentation). And he makes the Robinson Crusoe argument used by Witters: that it is _private_. If you think the bow tie is too incandescent, that is YOUR problem, since a sense datum is 'private'.
By 1980, we read in Dædalus: "From the point of view of strict empiricism, the attempt to go beyond sense data seems to fail."
where Dædalus's implicature does NOT seem to be that Empiricism, by holding this, fails.