Hacking, Chs. 17-19

The best idea in Chapter 17 is... that Nietzsche did not infer (or argue for) chance (148), but rather "he experienced it".  This suggests that chance, and the taming of it, has less to do with an argument against determinism, and more to do with a new positive and productive way of experiencing, seeing, &c..  But the chapter remained difficult for us at the end of discussion... What could it mean to say that "Necessity and chance are twinned" (148)?

The best idea in Chapter 18 is... that determinism of the twentieth-century micro-physics stripe is an invention of the 1870s (Cassirer)... but that there is another (older) kind of determinism that "excluded statistical law" (152).  It is this kind of determinism that the book tracks in its story of "the erosion of determinism".

[There was discussion, albeit tentative, about the taming of chance as emerging positively in its own right producing only as an after-effect the erosion of determinism, rather than the taming of chance as attempting to negate determinism.  Rather, statistics leaves the metaphysics of determinism-voluntarism (the question of free will) "outside our province" (Durkheim, quoted on p. 159).]

The best idea in Chapter 19 is... "Nothing is more commonplace than the distinction between fact and value. From the beginning of our language the word 'normal' has been dancing and prancing all over it" (163). The idea of the normal is both descriptive and prescriptive.

[And thus the normal, perhaps, is beyond the debate, or problematic, of determinism and voluntarism.]

Hacking, chs 14-16

The best idea in chapter 14 is... an explanation of the distinction between penchant (propensity, disposition) and determining cause (fatalism) (cf. 123).  This distinction helps resolve a worry about statistical fatalism. The worry arises because of a tension between a 'metaphysical' and a 'political issue (cf. 121).  The tension is that on the one side we have an enormous political success in altering social conditions to improve various aspects of life, health, crime, etc., and then on the other side this political success seems to undermine our standard sense of moral (or metaphysical) agency.  The distinction between propensity and determination helps resolve this tension as follows.  The political success of social reform operates on propensities of actors; it does not compel actors or necessitate anything; it is probabilistic.  This can be construed as leaving intact a metaphysics and morality of freedom; this is because social reform posits propensities not necessities.

The best idea in chapters 15 and 16 are... tbd.

Hacking, Taming, chs. 11-13

The most important idea in chs. 11 and 12 has to do with... the distinction between objective (frequency) probability and subjective (degree of belief) probability, as one entree to the gradual erosion of determinism.  See the chart below:

A second important idea in the discussion, though not necessarily in the chapters themseslves, concerns the erosion across the 19th century of, perhaps, determinism (represented by idealism) and essentialism (represented by persisting classical empiricisms).  Is the 'natural kinds' project then a companion to the 'probability and chance' project?

The most important idea in ch. 13 is ... the story of how Quetelet takes the idea of statistical mean and transforms it into a real quantity.  "This is a crucial step in the taming of chance. It began to turn statistical laws that were merely descriptive of large-scale regularities into laws of nature and society that dealt in underlying truths and causes."

Hacking, 'Taming', chs. 9-10

The best idea in Chapter 9 is... that, "A category of problems--pretty much what we now call 'social problems'--was created to be shared by joint experts, medical and legal" (76).


The best idea in Chapter 10 is... the idea of a distinction between statistics, statistical law, and statistical inference (10).  There is a rough chronology that first come statistics, then come laws of statistics, then come statistical inferences.  Is there also a conceptual/rational connection here?  Or is there a 'statistical style of reasoning' involving statistical inference that would not necessarily take the form of law?  Is the conclusion of every inductive inference necessarily a conclusion in the form of a law?

Hacking, 'Taming', chs. 6-8

The best idea in chapter 6 is... that of a methodological focus on practical problems: "This was not an abstract, intellectual event, but, as always in the taming of chance, a practical attack on an immediate and material problem" (p. 48).

The best idea in chapter 7 is... that sometimes we travel 'the road from scientific law to scientific measurement' (Kuhn) in the opposite direction, moving from numbers to law, which is heroized less by the grand theoretical scientist and more by the armies of counting bureaucrats (see p. 62).  The numbers give us a wealth of data for making "more inductions".  The 19th century search for constants in the numerical data, and found there 'laws' (which today we might call 'regularities' looking forward to p. 128).  The search for constants is epitomized in Babbage (cf. 58-9).  The years in which this searching stabilized was 1820 to 1840.  This was (for us) a platform leveraging the previously stabilized platform of bureaux of enumeration which had stabilized a little earlier, 1800s and 1810s.

The best idea in chapter 8 is... that, for Esquirol, "Madess... was English."  This is part of an account of how the very first self-consciously statistical laws were conceived in explanatory terms (p. 71); this was (as later chapters argue) not an easy combination.

Hacking, 'Taming', chs. 3-5

The best idea in chapter 3 is... that statistics was a chance encounter, namely that the solidification of statistical institutions was the process of a contingent intersection between 'public amateurs' and 'secret bureaucrats'.  Statistical bureaux later emerged when there came to be 'public bureaucrats', professionalization of amateur knowledge who were no longer doing their thing in secret.

The best idea in chapter 4 is... "We do not here want a history of institutions" (29).  But what is the difference between an institutional history and a historical epistemology (or conceptual history [not Koselleck])?  "Concepts are words in their sites. Their sites are sentences and institutions" (7).  The history of concepts involves the history of institutions, but the institutions are of interest as sites for the emergence of concepts.  Thus, "such an institution presupposes that there is a special type of knowledge" (29).

The best idea in chapter 5 is... the contrast between a deterministic conception of social law (in Prussia) versus a probabilistic conception of social law (in France and England).  One can have a deterministic or a probabilistic view of statistics (37).  One is apriori and the other is empirical -- one is deductive and the other inductive.  The stabilization of the latter, we suspect, is crucial to the history of how chance came to be taken quite seriously.  But more on this in later chapters...

Hacking, 'Taming' chs. 1 & 2

This term we are playing the game of "the best idea in this chapter is..."

The best idea in chapter 1 is ..."a seeming paradox: the more the indeterminism, the more the control".  The chapter sets up a story about how two processes proceed hand in hand: the erosion of determinism and the avalanche of printed numbers.  The printed numbers are an information that is a key vector in new techniques of control.  These printed numbers, this information and control, could be potent and valuable just insofar as determinism was eroded (or eroding).  An interesting aspect of this story, for us, is that the social sciences pave the way here, and the natural sciences almost seem to follow the avalanche enacted by the social -- this is interesting because it runs cross to a usual story in the history of science that natural science presses forward first.

In addition, a question that arose here concerned the methodological focus on "sentences" moreo than "institutions" (Hacking notes that he regrets this, 7) which perhaps goes hand in hand with a methodological interest in "styles of reasoning" rather than something that more explicitly invokes actions alongside reasons such as "dispositifs".  Perhaps this sort of methodological shift would enable a kind of inquiry that sees probability along a fifth dimension of politics (in addition to the dimensions of metaphysics, epistemology, logic, and ethics named on page 4).

The best idea in chapter 2 is that the erosion of determinism did not have to lead to the taming of chance.  Laplace (and even Hume) assumes a metaphysical necessity (Hume is just an epistemological skeptic about this, whereas Laplace an epistemological optimist).  Bichat (p. 14) denies necessity but preserves causality (via vitalism).  It was a further move to not only deny necessity but also strive to understand chance, and Charles Santiago Peirce (p. 11) is the figure or emblem for this.