Now that that's out of the way, I can explain why I make such extravagant recommendations.
A very English view of intellectual life, and country or even county English at that. All is tradition --- flexible, but traditional --- and implicit, unformalized, ``done'' or ``not done,'' at least the genuinely compelling stuff is. The merely ``magisterial'' authority of magistri is like the transitory & frowned upon statue law in relation to the common law of intellectual equity.
This is Tory phil. of sci., as almost all prior English-speaking phil. of sci. has been Whig --- excepting Kuhn, who, imitating an earlier and greater Tom, was Hobbesian. This has nothing, or almost nothing, to do with the political views and personal sympathies of the philosophers: Whehwell, for instance, was a Tory, and Coleridge's politics were simply disgraceful. Still, the Whig tone and attitude was there, maybe even that of Radicalism: progress, clarity, universal criteria, rationalization. Science at its best emerges as a well-oiled cognitive machine --- not necessarily or even usually for gradgrinding facts into laws, but still.
Toulmin, as an (ex-?) Wittgensteinian, is having none of it. He offers a generalized and universal defense of the paritcular local cognitive customs of intellectual disciplines. (This is the common problem afflicting all arguments for the particular in general; we shall see at the end how Toulmin attempts to heal this breach, and so close the argumentative circle.)
Rationality = fairness (quote Chuang Tzu paradox), != logicality (but it may be possible to formalize a lot more than he thinks). Conflating rationality with logicality, in his opinion, leads to either ignoring conceptual change, or to treating it relativistically. (There is a very amusing reductio ad absurdum of Kuhn in this connection.) I'm not sure how far this is actually so, though he's right that many of the ways in which people have tried to formalize rationality have been bad ones. He neglects (justifiably, perhaps) AI and cognitive science, but his description of, for instance, concepts (roughly, someone has mastered a concept if they've learned when to apply it, and what procedures of explanation or manipulation that application warrants) sounds rather like John Holland's classifier syustems.
Aggressively agnostic about whether or not _ideas in the head_ have anything to do with _knowledge_. He doesn't deny it, exactly, but he goes out of his way to phrase things in such a way that whether or not people have internal images or whatnot is neither here nor there.
A scientific discpline's ``transmit'' = goals of what to explain + current repertoire of concepts + idea of what an explanation should look like.
Problems = difference between what a discipline feels it should be able to explain, and what it currently can explain. (``The art of the soluble,'' as Medawar says.)
Constant production of conceptual variants, in the pursuit of getting those problems solved. Changes in explanatory ideals and disciplinary goals are rare and harder, justified essentially by either _what we've been able to do_ and _a bet about which direction will be progressive_, respectively. (Cf. Laudan's Science and Values, and his program of ``normative naturalism'' more generally.) Selection of these.
Speculations about human universals. Strangely, while he's fully aware of the neuropsychological evidence for highly specialized brain functions, and sees that this must have a (conventional) evolutionary explanation, plumps for something like a general talent which adapts itself to the regular features of human life, rather than evolved abilities. This is a merely empirical question, of course.
Inter-relation of social structure and intellectual organization. How the trick is turned, of making sure that usually and on balance ``magisterial'' authority of eminent scientists coincides with the ``disciplinary'' authority of what people should actually think.
Closing the circle. Justifying things overall by by appealling to the perspective of the entire past. I'm really not sure that this works; for, even within the sciences, or natural philosophy, there have been such radically different ideas entertained as to what their proper goals are, that it seems an act of pure optimism to suppose any sort of rational adjudication could be reached on this score.