The social construction of What?


Ian Hacking





I.    Introduction

A.  In this chapter, Hacking lays out several fundamental disagreements about the natural sciences that are at the heart of the “science wars”

1.  he calls them “sticking points”

2.  they range from the philosophical to the nearly political

a.  social scientists may say that the science wars are about challenging a certain image of the scientist and may prefer to start there

b.  however, Hacking thinks we can gain greater clarity things by starting with the philosophical issues

c.  the philosophical differences concern very old issues that have been made to seem contemporary by using the language of “social construction”

B.  What are the natural sciences? (64)

1.  represented by physics, chemistry, molecular biology

2.  it is here that social constructionism meets the greatest resistance

3.  nobody is surprised to learn that, say, primatology reflects the cultural values of the scientists who study apes

C.  Who are the social constructionists about science?

1.  Hacking’s two exemplars of social construction are two authors who have already been mentioned:

a.  Pickering’s (1984) Constructing Quarks

b.  Latour and Woolgar’s (1979) Laboratory Life

2.  he chooses these even though they are old books by now because:

a.  they actually use the term (66)

b.  the authors are still active

c.  their descriptions of laboratory life were thought to be faithful by scientists, even when they rejected their philosophical conclusions

d.  some scientists regard them as the enemy

3.  there are many others who practice the various science studies disciplines that are loosely called “constructionists” including the Edinburgh and Bath schools, etc. (65)

4. but these do not serve as such good examples of constructionism

a.  for example, the Strong Programme is not quite as connected to constructionism as people seem to think

b.  they are more important for the third sticking point, external explanations of the stability of science

D.  Distinctions (66)

1.  Hacking would agree that science is a social activity but only after making a distinction

2.  Science may be understood as either:

a.  scientific activity, in which it is trivially true that it is a social activity

b.  an assemblage of truths or even of testable hypotheses, which is not a social activity

3.  sociologists are more interested in the process, scientists in the product (67, q.v.)

4.  many people don’t like making distinctions

5.  Hacking, however, defends the practice

6.  First, he discusses an example of the sort of misunderstanding that occurs when one does not:

a.  example: (67)

1.) Dawkins’s claim that nobody is a social constructionist at 30,000 feet

2.) and Collins’s reply that Dawkins has money in his pocket in the sky and money is a social construction

b.  Hacking replies

1.) that nobody denies that things that depend on social things are social

2.) although many constructionists dislike the sciences, they are not saying

a.) that what science says is not true

b.) or that the artifacts whose design depends on science do not work

7.  Second, he argues that sometimes making a distinction can end a controversy by showing that the opponents were really talking about different things (68)

8.  However, the science wars are not so easily resolved as the parties to the dispute have fundamentally different philosophical points of view

II.  Sticking point # 1:  Contingency (68)

A.  Pickering's Constructing Quarks

1.  as Hacking mentioned in ch. 1, Pickering's claim is not that the object or even merely the idea of quarks is constructed (68-69)

2.  nor is he denying that quarks exist (70)

3.  rather, his point is that the development of the theory of the quark in high-energy physics was not inevitable (69)

a.  funding for physics could have just ended

b.  or, physics could have developed in a non-quarky but equally successful way; perhaps the old pre-quark physics could have just continued

B.  Success

1.  what could it mean to have an equally successful but nonquarky physics?

2.  This raises the whole issue of whether there are standards of success that are not simply internal to the science -- the "paradigm" -- itself

3.  Lakatos provides a neutral way of evaluating success (70)

a.  a research program is progressive if it continues to make true predictions, etc. (q.v.)

b.  Hacking is not proposing that Lakatos's philosophy is the correct one (70)

c.  Rather, he is simply saying that it gives us one way to understand how you could have more than one equally successful physics

C.  Pickering's claim could be generalized beyond quarks to other theories in science (p. 70, q.v., for list)

1.  none of these theories were inevitable

2.  most scientists regard this claim as ridiculous

3.  the sticking point here is not really about truth or reality

4.  indeed, Pickering avoids the term "truth" and speaks of "resistance" and "accommodation" instead

D.  Resistance and Accommodation (71)

1.  resistance

a.  in experimental work, the apparatus does not always behave as expected

b.  as Pickering puts it, the world "resists"

2.  scientists then have to "accommodate" themselves to that resistance by:

a.  revising the theory

b.  revising beliefs about how apparatus works

c.  modifying apparatus itself

3. the result of doing one or more of these things is a "robust fit" among all these elements (71)

E.  Robust Fit

1.  Pickering's view of experimentation is like that of Duhem

2.  That is, when results don't agree with expectations, we can revise either the theory or one of the auxiliary hypotheses

3.  What Pickering adds is that we may also modify the equipment itself (72)

4.  In modern physics, Hacking adds, we may also achieve a robust fit by modifying the phenomenology

a.  Duhem distinguished the concrete instrument from the schematic model of it, expressed in symbols with the aid of theory

b.  The phenomenology is an account or interpretation of experimental results (72)

F.  Contingency means no predetermination

1.  to sum up Pickering's position, there could have been an alternative successful research program (q.v.)

a.  with a different theory, phenomenology, apparatus, etc.

b.  and a different series of robust fits among these elements (72)

2. what Pickering's constructionist thesis amounts to is that before a robust fit has been achieved, it is not determined what it will be (73)

G.  Contingency does not mean underdetermination

1.  Pickering's thesis is not the Quinean one that our choice of theory is underdetermined by experience

2.  It's not just a matter of choosing a theory but of meddling with theory, apparatus, etc. (q.v.)

3.  What changes we make is not predetermined – he doesn't even seem to accept that there are constraints on what changes we can make

4.  Hence, when Pickering says that quarks were not inevitable, he does not mean that physicists could have chosen a different theory (73-74)

5.  Rather, he means that there is more than one way the physicists could have adapted or accommodated to resistance (74)

6.  Here Hacking offers an analogy with biological evolution -- no set of conditions in the world pre-determines the path that evolution must take (q.v.)

H.  Alien science (74)

1.  many physicists find it inconceivable that physics in the 1970s could have taken something other than the quark road

a.  they reject Pickering’s claim that the old physics could have continued

b.  the structure of physics and even of the apparatus would have had to be the same, even if the names were different

c.  they hold that any successful physics would have to be equivalent to the physics we have

2.  But what does equivalent mean?

3.  Weinberg:  if we discover some intelligent aliens and translate their science, we will find that they have discovered the same laws (75, q.v.)

4.  this focus on translation, however, raises some philosophical problems:

a.  as Davidson would argue, in order to translate their language, we would need to assume that they shared many of our beliefs in the first place (75, q.v.)

b.  similarly, Quine would argue that we would say that the alien was talking physics only if we could translate what he was saying into something we recognize as physics (75, q.v.)

c.  although Hacking may not agree with this use of these philosophers, he does not see any other way to clarify what Weinberg means by equivalence

5.  Weinberg also suggests that Maxwell’s equations should be deducible from any sound physics

6.  Hacking has 3 problems with this suggestion, as well:

a.  suppose computers had been more successfully developed by 1850 (76)

1.) then the analytic mathematics of Maxwell’s equations would have been unnecessary and we could have bypassed it

2.) a physicist might object that the structure of the computations would have conformed to Maxwell’s equations

3.) but Hacking argues that the notion of “conforming” is even more obscure than that of “equivalence”

b.  a bigger problem with Weinberg’s suggestion is that obtaining the deduction in question is not a trivial exercise

1.) to deduce the consequences of Newton’s mechanics, Laplace and Lagrange had to invent new branches of mathematics (76)

2.) Hacking appears to be suggesting that deducing Maxwell’s equations from some alien physics could be like this

3.)  deducibility is not transparent

c.  even if we have two or more formulations of a theory that are scientifically or experimentally equivalent, they may have different uses or function (76-77, q.v.)

1.) that is, they may be different from a psychological point of view, from the point of view of how useful or helpful we find them (77)

2.) the choice between alternative formulations is not arbitrary

I.  Convergence

1.  philosophers have been a little more cautious than scientists in trying to formulate what the philosopher Bernard Williams calls “the absolute conception of the world”

2.  according to Williams, science, unlike ethics, will converge on an answer, and the best explanation of that convergence is that the answer represents how the world is

3.  Hacking questions this suggestion, too:

a.  small-scale convergence

1.) this happens when we get a “robust fit” among theory, apparatus, etc. and it happens all the time (77-78)

2.) but the best explanation of such robust fits is not that that’s how the world is:  the world does not predetermine which robust fit is achieved (78)

b.  big-scale convergence

1.) that is, on the big picture, not specific answers to specific questions

2.) this could mean that science will converge on an answer or on the one and only answer

3.) if the former, convergence is consistent with contingency

c.  unique-ultimate

1.) that science will converge on not just an answer but on only one

2.) but even that sort of convergence is consistent with the contingency thesis, since there could be more than one road to that answer (78)

J.  The Sticking point

1.  the constructionist maintains that the development of science is contingent

2.  in the case of physics, the constructionist says:

a.  there could have been an alternative to the theory of quarks that was just as successful as current physics (78-79)

b.  this alternative physics would have in no sense been equivalent to our physics (79)

3.  physicists tend to deny this

a.  they would demand that one actually show them an alternative

b.  their example of what could not be different may not be quarks but Maxwell's equations, the second law of thermodynamics, or the velocity of light

c.  when Hacking calls them inevitablists

1.) he does not mean that they think physics was simply inevitable

2.) they mean that if physics were successful, it would inevitably have had to have happened our way

K.  Metaphysics (79)

1.  strictly speaking, the contingency thesis is compatible with any metaphysics

a.  in particular, it's consistent with scientific realism (79-80)

1.)  the view that science aims at the truth

2.)  and that the entities it refers to exist

b.  it's also possible that an anti-realist might reject the contingency thesis (80)

2.  more loosely speaking, however, there is a metaphysical issue at stake here (79)

a.  someone may feel strongly that we don't live in the sort of world where the contingency thesis could be true

b.  but that's a non-empirical, non-testable, and hence metaphysical claim

III.  Sticking point # 2:  Nominalism (80)

A.  Facts

1.  “fact” is one of Hacking’s “elevator words”

a.  like true, real, etc.

b.  these are not words for things that are in the world

c.  rather, they are words used to say things about what we say about the world

2.  Latour and Woolgar’s Laboratory Life

a.  facts are made (81)

b.  indeed the word comes from Latin facere, to do or to make

c.  they do not deny that facts or reality exists

d.  rather, they wish to say:

1.) that facts are the result, not the cause of scientific work (q.v.)

2.) we cannot explain why something becomes a fact by appealing to reality

3.  Philosophers like Hacking, on the other hand, are uncomfortable with the idea of something becoming a fact – if it’s a fact, it’s always a fact

4.  Nevertheless, he thinks Latour and Woolgar are correct in saying that the truth of an idea should play no role in explaining why anybody holds it (81-82)

a.  e.g., in explaining why someone believes in the big bang, you should list all the reasons.  To these reasons, you should not add “and it’s true.” (81, q.v.)

b.  this point is not special to social construction – it’s a general point about explanation

B.  Nominalism (82)

1.  for Hacking, the issue at stake here in the science wars resembles the old philosophical debates about universals between the nominalists and the realists

a.  Nominalism is the philosophy that there is nothing particular to a group of things that share a general name except that they share that name (82-83)

b.  The realists say that good names will “carve nature at the joints,” while the nominalists deny that nature even has joints (83)

2.  in its more contemporary version,

a.  one side hopes that the world may be structured in the way we describe it

b.  Hacking prefers to call this side “inherent structurism” because “realism” has taken on other meanings

c.  the other side says that all the structure that we conceive lies within our way of representing the world (83, q.v.)

C.  The Sticking point (84)

1.  nominalists are a kind of radical empiricist

2.  they are more radical than the anti-realists who are concerned about unobservable entities (84)

3.  the nominalists are not concerned with the problem of observability, but are concerned with the more general problem of kinds of things even in the observable world

IV.  Sticking Point # 3:  Explanations of Stability

A.  Understanding stability

1.  to many, it seems that a lot of science is here to stay (85)

2.  examples that typically occur in debate include:

a.  Maxwell’s equations

b.  Second law of thermodynamics

3.  perhaps because philosophers like Kuhn and Popper lived in the 20th century when physics underwent major changes, science seemed unstable to them and the problem was to understand revolution

4.  perhaps now the problem is to understand stability

5.  to say that Maxwell’s equations are a stable part of science is not necessarily to say they are exactly and universally valid:  see quotation A from Steven Weinberg, p. 86

B.  Culture and Science (86)

1.  Maxwell’s equations (86)

a. historians like Norton Wise have argued that culture and science are inseparable

1.) for instance, he argues that Maxwell’s equations came from the work of deeply religious physicists

2.) other scholars emphasize the role of empire and the laying of telegraph cables

b. the physicist Weinberg, on the other hand, replied that whatever cultural influences were there in the development of these laws, they have since been refined away

2.  Second Law of Thermodynamics (87)

a.  a constructionist might point out that it takes its name from the old name of the steam engine, that it still uses the concept of work, betraying its origin in the industrial revolution

b.  but a scientist would argue that the content of the law, what it means, is independent of its history

3.  Hacking tends to side with the scientists

a.  although one turns to history to find the questions that led to these laws, the content of these laws and equations are free of history (87)

b.  perhaps, however, if not the content, then the form that we have given to this knowledge is historically determined – he takes up this suggestion in ch. 6 (87-89)

C.  A Big Jump (88)

1.  although quotation A from Weinberg (p. 86) seems  uncontroversial (so far), it led him immediately to quotation B on p. 88, which is controversial

a.  quotation A did not seem to touch on the first two sticking points, of inevitability and nominalism

b.  but B, with phrases like “as real as anything else we know,” suggests stands on both

1.) it suggests that they are irresistable, inexorable – hence inevitable

2.) Weinberg even says there are no alternatives to Maxwell’s equations (89)

3.) also, Weinberg seems to think that these equations reflect the way things are – the inherent structure thesis

2.  Hacking adds that even A would be controversial to Kuhn

a.  Kuhn insisted on the necessity of revolutions (89)

b.  While Weinberg emphasizes stability

c.  Hacking compares them to two men looking down opposite ends of a telescope (q.v.)

d.  Also, Kuhn, in denying that there is progress toward a true representation of reality, was a nominalist (90)

D.  External Explanations (90)

1.  scholars in the science studies disciplines have offered different sorts of explanations for science that appeal to external factors

2.  Edinburgh school

a.  interests

b.  don’t take construction metaphor literally

c.  but do emphasize the social (91)

3.  Latour appeals to a network of agents and events (90)

a.  if you doubt something in science, there are endless things that it's connected to, including people, that you could challenge

b.  actually does not choose to separate the social from the natural (91)

4.  When Weinberg in quotation B complains that Latour denies the "objective nature" of science, part of what is really bothering him is Latour's willingness to appeal to external factors (91)

a.  it does no good for Latour and his collaborators to say that he does not deny reality, the facts, etc.

b.  Weinberg thinks external factors are irrelevant to science, Latour thinks they are relevant

E.  Rationalism and Empiricism (91)

1.  Hacking sees an analogy between this current debate and the traditional debate between rationalists and empiricists in philosophy

a.  rationalists sought good reasons internal to our knowledge for what we take to be true

b.  empiricists turn to something external:  sense experience

2.  Hacking and Alan Nelson describe the current debate as one between rationalists and constructivists

a.  rationalists hold that there are good experimental or theoretical reasons for the stability science achieves

b.  for constructionists, these reasons are not decisive – something else settles the dispute (91-92)

F.  The Sticking Point (92)

1.  the constructivist says explanations of stability appeal to external factors

2.  their opponents say not

V.    Anti-Authority by unmasking

A.  Hacking's three sticking points all concern intellectual or philosophical matters

B.  Another way to look at the science wars is from the emotional side, at what's "really bothering" the scientists

1.  cf. the quotation from Nelkin on p. 63

2.  scientists, at least high-energy physicists, feel their authority is being challenged (93)

a.  after World War II, high-energy or particle physics was the queen of the sciences (93)

b.  now, people in this field have trouble finding jobs

c.  some would argue that this is why they are kicking up a fuss about social construction

3.  Hacking, however, would argue that this makes for a bad argument (93, q.v.)

a.  even if it were true that these scientists were upset, that their authority is being challenged, etc.

b.  that does not touch the question as to whether scientists' arguments against the constructionists are well-founded

C.  Unmasking scientists (94)

1.  referring back to his 6 grades of constructionism on p. 19, he finds that his favorite examples of constructionists -- Pickering, Latour, Woolgar -- are not reformist, etc. but ironist

a.  that is, sees that what we have in science is highly contingent, the product of social factors

b.  yet is content to leave things the way they are

2.  nevertheless, Hacking also finds a strong element of unmasking in many constructionists

a.  their target is not the propositions that scientists hold true

b.  but the image that scientists want to project of themselves

D. The ideology of science

1. as he explained in ch. 2, unmasking is not refuting ideas but showing the function they serve (94, q.v.)

2. constructionists think there is an extra-theoretical function for the inevitabilism, inherent-structurism, and rejection of externalism (q.v.)

a.  they are part of an ideology of science that preserves the authority of scientists (94-95)

b.  what they are unmasking is the physicists' claims to be probing the deepest truths about reality

3. hence the sticking points (95)

a.  as we've seen, there is a strong metaphysical element to these sticking points

b.  but metaphysics has always been a sure way to bolster authority – e.g., the divine right of kings

c.  by unmasking the metaphysics, constructionists are undermining scientists' authority

VI.  Left and Right politics (95)

A.  What’s curious about the science wars is that both sides consider themselves as representing the left

1.  scientists

a.  Sokal, for instance, considers himself on the left and is dismayed to see the other side claiming to represent the left (95)

b.  Scientists see themselves as defenders of both the oppressed and objective knowledge

c.  Indeed, objectivity, truth, fact, etc. are regarded as the last defense of the weak and oppressed (96)

2.  constructionists are on the left insofar as they are engaged in unmasking the established order (95)

B.  Feminists, on the other hand, may reject both left/right politics and objectivity as part of the established male order – the ideal of objectivity having been something used against women, who have been regarded as subjective (96)

C.  Hacking finds himself torn between the sides of this debate

VII.  Kuhn and Feyerabend (96)

A.  despite Feyerabend’s reputation as an intellectual anarchist, Hacking finds Kuhn to be more of a constructionist

B.  Kuhn as constructionist (97)

1.  although he never talked about social construction, he fits Hacking’s checklist of 3 sticking points (97)

2.  contingency

a.  he held that progress in science is away from past science, not toward the truth about the world, which is a contingency thesis

b.  although normal science follows a pre-ordained path, the result of a revolution, the new paradigm, is entirely contingent

3.  Kuhn is also a nominalist

4.  Explanations of stability

a.  as Hacking mentioned earlier, Kuhn was dubious as to whether there was any stability

b.  but Kuhn was not opposed to offering external explanations

C.  Kuhn’s book also undermined the ideology and the authority of science in quite an unusual way

1.  on the one hand, most scientists were portrayed as mere puzzle-solvers, not seekers after deep truths about the universe

2.  and when there was a revolution, this had less to do with reason and evidence than with the old-timers dying off (97-98)

D.  Feyerabend (98)

1.  was opposed to the authority of science but not in any social constructionist or unmasking way

2.  contingency thesis

a.  he never defended this

b.  he was a methodological pluralist, but this is not the same thing:  there could be many methodologies, with each developing in an inevitable way

3.  nor did he give external explanations of stability

4.  the only sticking point where he agrees with the constructionist is his nominalism

E.  conclusion (98-99)

1.  Kuhn could be called a social constructionist

2.  But not Feyerabend – indeed he would have mocked it as a new orthodoxy (99)

VIII. Check list (99)

A.  Hacking gives Kuhn all 5’s on the three sticking points

B.  As for himself, he’s ambivalent (q.v.)