first time i heard about tukey was by reading the papers of shannon, who gave him credits for the word bit.

The choice of a logarithmic base corresponds to the choice of a unit for measuring information. If the base 2 is used the resulting units may be called binary digits, or more briefly bits, a word suggested by J. W. Tukey. A device with two stable positions, such as a relay or a flip-flop circuit, can store one bit of information.(a mathematical theory of communication)

others say, that these bits actually don't mean the same thing

But the Tukey bit and the Shannon bit are not the same thing. The popular notion of a bit is simply a unit of storage, with a 0 or 1, and Tukey coined this as short for binary digit. The Shannon bit is a unit of information, and a storage bit contains at most one bit of information, but quite often a lot less. (shannon documentary)

anyhow, the guy has not just an interesting biography, but also some lines that too often are ignored. he worked togther with shannon and richard feynman, with whom he discovered that things like counting is different for different people. they played around with things like: is it easier to keep track of time simultaneously while speaking or while reading; and it seems they had fun with things alike ...

notice that the question comes first and the probability models last. (how davies's data sets might reasonable be approached)

davies has gone a long way toward procedure orientation. her discussion at the top ... is summarized by the sentence:" in other words models are chosen to produce [as properties of the procedures they 'prescribe'] desireable operational characteristics."

this is a long step forward, models are now valued for their (formal) consequences, rather than for their truth. This is good, but does not go far enough.

Since the formal consequences are consequences of the truth of the model, once we have ceasd to give a model's truth a special role, we cannot allow it to 'prescribe' a procedure ... in short, we need to change for assumption-orientation to procedure-orientation. (issues relevant to an honest account of data-based inference)

three articles by j.w.tukey