In praise of the inexact, the inelegant and the unassuming

A standard philosopher attending my recent talks would be very disapproving: ‘You are seriously inconsistent!’ See the title of this talk: ‘In praise of the inexact, the inelegant and the unassuming’. But look also at the theme for many of my other talks, taken from the WWII slogan: ‘Loose talks costs lives.’ Here I aim to reconcile these views. I will defend my on-going attacks. Loose talk is bad -- on its own. But, I will argue, loose talk is great when it has an appropriate tangle of extra work behind it. I will illustrate using examples of different ways to report the ‘same’ scientific results. My ideas build heavily on JL Austins’ insight that we do things with words and I apply Mary Hesse’s accounts of models as analogies to the reports of research results, which are inevitably loose. Loose talk – importantly: given the tangle of work behind it – insinuates a positive, negative and neutral analogy. When it is read against a good tangle of background work, loose talk, like a Hesse model, tells you what you can and cannot do with it.