1953 And all that.A tale of two sciences

Philip Kitcher, 1984

Presented by: Yoav Francis, 15/01/2017

Philip Kitcher• Born 1947. American philosophy professor

who specializes in the Philosophy of science, of biology and of mathematics and literature• Phil. Professor at University of Columbia • Fun fact: was a supervisor of Peter Godfrey-

Smith

Agenda• Background and preliminaries• Classical/molecular genetics, Mendel’s laws, laws in biology, genes, reduction

• Kitcher (1984)• The question and goals of his inquiry• Requirements from an intertheory reduction: R1, R2 and R3

• Refutation of reduction of classical to molecular genetics• Alternative structure for classical genetics• Reanalyzing the relation between classical and molecular genetics• Positions for a reductionist and an anti-reductionist approach in theories

• Conclusions and questions

Timeline and relevant thinkers • Mendel's laws – 1850

• Rediscovery of Mendel’s laws – 1900• Morgan – Genes are (mostly) chromosomal segments - 1910• Watson and Crick – Model of the DNA - 1953• Kitcher’s paper- 1984

Preliminaries• Mendel's laws • First and second laws• A word about laws in Biology

• Classical genetics• Molecular genetics• The concept of “genes” in classical and molecular genetics• Reduction

Mendel's laws• First law – law of segregation• Every individual organism contains two alleles for each trait, and these alleles

segregate (separate) during meiosis such that each gamete contains only one of the alleles.

Mendel's laws• Second law - law of independent assortment• The Law of independent assortment states that

alleles for separate traits are passed independently of one another from parents to offspring. That is, the biological selection of an allele for one trait has nothing to do with the selection of an allele for any other trait.

A word about laws• Are Mendel’s “laws” really laws? What are laws in biology?• Usually in science we’d look for “laws of nature” – laws which are

spatiotemporal, and have a sort of necessity rather than just describe how things merely happen to be.• Is Mendel’s first “law” a law or regularity/pattern? • On the one hand - It won’t necessarily be true in the future (nor was in the

past). Hench it can’t be defined as a “law of nature”.• On the other – It’s not an accidental regularity.

• “Mendel’s first law is a predictable result of the operation of mechanisms the are contingent historical products” (PGS, 2014, p. 13)

A word about laws• Biological patterns show a degree of resilience (PGS, chap. 2)• A resilient pattern holds across many cases, and gives us reason to believe it will

hold in other possible (non-actual) cases.• A regularity has more or less resilience – not a yes/no question with single scale of

comparison (Mendel's law has some resilience)• “Many sciences use laws that are clearly recognized as approximation” (Kitcher,

p.342)• Does this view (where laws are patterns with some/a lot of resilience) apply only

to Biology?• Sandra Mitchell (2000) – No, applies to physics and all sciences• Alternatively, Physics can be viewed a special case since it describes laws that govern the

fundamental working of the world (PGS, chapter 2).• It’s the working of physics laws in organisms that give rise to patterns that are not exactly like

physical laws – but have some varying level of resilience. (PGS)

Classical genetics• Classical genetics stems from the studies of T.H.Morgan and is the

successful outgrowth of the Mendelian theory of heredity• “Philosophers often identify theories as small sets of general laws”

(Kitcher, p. 340) – what are the laws of classical genetics?• “It is much easier to identify classical genetics by referring to the subject

matter and to the methods of investigation, than it is to provide a few sentences that encapsulate the content of the theory.” (ibid.)• “We find it hard to pick out any laws about genes” (ibid.)

Classical genetics• Perhaps Mendel’s laws are the base laws of classical genetics?• “Once it was recognized that genes are (mostly) chromosomal segments…we

understand that the (second) law will not hold in general” (Kitcher, p. 342)• Even if we amend the law so it is general, it will still not work:

“There can be interference with normal cytological processes so that segregation of nonhomologous chromosomes need not be independent” (ibid, p. 342)

• “Mendel’s second law…becomes irrelevant to subsequent research in classical genetics” (ibid.)• “What figures largely in genetics after Morgan is the technique” (p.343)

Genes• At the base of classical genetics is the concept of a gene, the

hereditary factor tied to a particular simple feature (or character).• By the end of the period of classical genetics, genes were seen as

unknown physical entities arranged on chromosomes, passed on in the ways described by Mendel, and each responsible for one Enzyme (PGS, Chap. 6)

Molecular genetics• Began with Watson and Crick in 1953 with the discovery of the DNA. • Principles:• “Most genes are segments of DNA” (Kitcher, p.343)• “We know the molecular structure of DNA” (ibid)• “There is no general law governing the distribution of all segments of DNA

and RNA” (ibid., p.346)• “It is becoming increasingly obvious that genes are not always transcribed,

but play a variety of roles in the economy of the cell” (ibid., p.345)

Genes - Molecular vs. classical genetics

• “It seemed for a time that a single, definite kind of thing played a certain role – and that unity partly dissolved” (PGS, p.85)• “To some extent, the classical gene had its nature explained”• “To some extent, the concept has been augmented”• “To some extent, it has been replaced”

• Thought experiment: “Suppose we start from scratch, with the finer- grained information we have now, and had not gone via Mendel and Morgan. Would we talk about genes as units at all?” (ibid)• “What cells contain is genetic material. And different stretches and chunks of it

play different roles.• Seems like postulation of Mendelian “factors” followed by Morgan’s “particle-

like” genes was indeed productive, but today things are analyzed differently (ibid)

Reduction• ”one of the most used and abused terms in the philosophical lexicon”

(The Oxford Companion to Philosophy)• Methodological reductionism – the scientific attempt to provide

explanation of a system in terms of the individual parts and their interactions• Such a reductionism does not imply that the whole system is “nothing but” it’s

lower level parts.• The entire system can have features that none of the parts have• Reducible properties – can be explained at a lower level• Emergent properties – cannot be explained at a lower level, thus irreducible.

• In phil. of mind, consciousness is sometimes said to be an emergent property• “It has material basis in the brain – we know what the neurons are doing, yet we can’t

explain how these processes give rise to consciousness” (PGS, chapter 2)

Theory reduction (Nagelian Reduction)

“A reduction is effected when the experimental laws of the secondary science (and if it has an adequate theory, its theory as well) are shown to be the logical consequences of the theoretical assumptions (inclusive of the coordinating definitions) of the primary science.” (Nagel 1961: 352)

“The Nagel model of reduction thus seems to suggest that a reduced theory is directly explained by the corresponding reducing theory… in contrast with indirect reductions—those that hold in virtue of the fact that the reducing theory explains the occurrence of the phenomena of the reduced theory (rather than the reduced theory itself)” (Stanford enc. of phil., “Scientific Reduction”)

Theory reduction (Nagelian Reduction)

“Scientific theories are regarded as sets of statements. To reduce a theory T2 to a theory T1, is to deduce the statements of T2 from the statements of T1” (Kitcher, p. 338)

𝑇 2>𝑇1Reducing theoryReduced theory

(Molecular genetics)(Classical genetics)

Theory reduction (Nagelian Reduction)

“We are allowed to supplement the statements of T1 with some extra premises connecting the vocabulary of T1 with the distinctive vocabulary of T2 (so called bridge principles)” (Kitcher, p. 338)

𝑇 2>𝑇1Reducing theoryReduced theory

“…this condition is dangerously vague” (p.338)

Kitcher’s question(s) and goal• Goal: “My plan in this paper is to offer a different perspective on

intertheoretical relations” (p. 337)• “What is the relation between theory of classical genetics and the theory of

molecular genetics? How does the molecular theory illuminate the classical theory?” (p. 336)• What shall classical genetics be used for following the introduction of

molecular genetics?• Muller’s hope: “Must we geneticists become bacteriologists, chemists and

physicists, simultaneously with being zoologists and botanists? Let us hope so” (1922)• Kitcher believes Muller’s hope was positively answered. But how so?

Refutation of reduction of classical to molecular genetics

R1 is false• Why is R1 required? • “…We are going to need to find general principles, identifiable as the central

laws of classical genetics, which can serve as the conclusions of reductive derivations”

• Why is R1 false?• We’ve presented that general laws in classical genetics are not to be found (in

particular Mendel’s laws do not serve as such general laws)

R2 is false• Why is R2 required? • “…If we are to derive such laws from molecular biology, then there must be

bridge principles connecting the distinctive vocabulary figuring in the laws of gene transmission… with the vocabulary of molecular biology” (p. 340)

• Why is R2 false?• Kitcher assumes to the contrary that R1 is true – laws do exist in classical

genetics, and he assumes that they are Mendel's laws.• A bridge law will be of the form:

R2 is false (cont.)• Why is R2 false?• Since (most) genes are segments of DNA (as discussed), “The problem of

providing a statement of the above form would be to say, in molecular terms, which segments of the DNA count as genes”• “The criterion is not general – not every gene is transcribed on mRNA” (p.344)• Kitcher fails to find such a bridge principle.

• Possible objections• A brute-force enumeration of all possible Mx (possible because we have a

finite number of organisms) in the form of disjunction (V/”or”) over all DNA segments that are genes would give us the bridge laws

R2 is false (cont.)• Kitcher claims that such a brute force approach also fails:

• “I claim that more than this is needed to reduce a law about gene transmission. We envisage laws as sustaining counterfactuals, as applying to examples that might have been but which did not actually arise. To reduce the law it is necessary to show how possible but nonactual genes would have satisfied it. Nor can we achieve the reductionist's goal by adding further disjuncts to the envisaged bridge principle. For although there are only finitely many actual genes, there are indefinitely many genes which might have arisen.” (p. 345)

• An attempt to find a weaker version of the form of the bridge law also fails (p.346)

R3 is false• Why is R3 required?

• “If the derivations are to achieve the goal of intertheoretical reduction then they must explain the laws of gene transmission” (p. 340)• “The major goal of reduction” (p.347)

• Why is R3 false?• Kitcher assumes to the contrary that R1 and R2 are true – there are laws in

classical genetics and there are bridge principles between molecular and classical genetics.• “In explaining a scientific law, L, one often provides a deduction of L from other

principles. Sometimes it is possible to explain some of the principles used in the deduction by deducing them, in turn, from further laws. Recognizing the possibility of a sequence of deductions tempts us to suppose that we could produce a better explanation of L by combining them, producing a more elaborate derivation in the language of our ultimate premises. But this is incorrect.” (p. 348)

R3 is false (cont.) – PS Processes• The molecular derivation forfeits something important• PS-Process: a process in which homologous chromosomes are

separated by a force so that one member of each pair is assigned to a descendant gamete. This is the original cytological explanation. (349)• “PS-processes are heterogenous from the molecular point of view.

There are no constraints on the molecular structures of the entities which are paired or on the ways in which the fundamental forces combine to pair them and to separate them. “ (p. 349)• “the explanatory power of the cytological account can be preserved

only if we can identify PS-processes as a natural kind in molecular terms” (ibid)

R3 is false (cont.) – Multiple realizability

• “PS processes are realized in a motley of molecular ways” (p. 350)• “The molecular account objectively fails to explain because it cannot bring out

that feature of the situation which is highlighted in the cytological story” (ibid)• “appeal to molecular biology would not deepen our understanding of the

transmission law. Imagine a successful derivation of the law from principles of chemistry and a bridge principle of the form (*). In charting the details of the molecular rearrangements the derivation would only blur the outline of a simple cytological story, adding a welter of irrelevant detail. Genes on nonhomologous chromosomes assort independently because nonhomologous chromosomes are transmitted independently at meiosis, and, so long as we recognize this, we do not need to know what the chromosomes are made of.” (p. 347)

Is reduction in physics multiple-realizable too?Possible answer

• In support of the “Local Reduction Move”, Enc (1983) has drawn an analogy with thermodynamics (cf. Churchland 1986 and Churchland 1988). Heat, he argues, is multiply realized at the level of microphysical interactions. Temperature-in-gases is different from temperature-in-solids, which is different from temperature-in-plasmas and temperature-in-a-vacuum. The multiple realizability of heat, however, does not imply that thermodynamics has not been reduced to statistical mechanics; it merely implies that the reduction proceeds piecemeal. Temperature-in-gases is identified with one type of mechanical property; temperature-in-plasmas, with a different mechanical property, and so on. Thermodynamics is thus reduced to statistical mechanics one lower-level domain at a time through the mediation of restricted domain-specific thermodynamic types: temperature-in-gases, temperature-in-solids, and the like. Something similar could be true of psychophysical reduction. Psychology could reduce to physical theory by way of various domain-specific mental types such as pain-in-humans and pain-in-Martians. (Internet encyclopedia of philosophy)

Root of the trouble• The root issue is the failure of R1 – the fact that classical genetics has

no laws (p. 351)• Classical genetics doesn’t have laws that act as “set of statements” we

try to axiomize (ibid.)• “The statements advanced by the great classical geneticists seem

more like illustrations of the theory than components of it.” (ibid.)• Only alternative is: “that there are generals laws in genetics, never enunciated

by geneticists but reconstructible by philosophers”

Kitcher’s alternative view for classical genetics

• Since we cannot define classical genetics as a theory with a set of laws, Kitcher gives a definition of it as a “scientific practice” (p. 352)• A practice has (ibid.):

• “a common language used to talk about the hereditary phenomena”• “a set of accepted statements in that language” (corpus of beliefs about inheritance)• “a set of questions taken to be the appropriate questions to ask about hereditary

phenomena”• “a set of patterns of reasoning which are instantiated in answering some of the

accepted questions”

• “The practice of classical genetics at a time is completely specified by identifying each of the components just listed” (ibid)

How does classical genetics being defined as practice?

• “The initial family from which the field began is the family of pedigree problems” (p. 354)• “Arise when we confront several generations of organisms.”• “The questions that arise may be to understand the given distribution of

phenotype”• “classical genetic theory answers such questions by making hypotheses about

the relevant genes and their distribution”• Each case of a pedigree problem can be characterized by a set of:• Data – set of statements describing the distribution of phenotypes among

organisms in a particular pedigree.• Constraints – general cytological information. E.g.. The thesis that genes are

chromosomal segments, and the principles that govern meiosis.• A Question – such as “what is the expected distribution in such and such case”

How to explain different versions of classical genetics?

“Each version of classical genetic theory contains a pattern for solving pedigree problems with a method for computing expected genotypes which is adjusted to reflect the particular form of the genetic hypothesis that it sanctions” (p.357)

“We can take each version of classical genetic theory to be associated with a set of conditions…which govern admissible genetic hypotheses.”

“…The patterns of reasoning used to resolve cases of the pedigree problem are constantly fine-tuned as geneticists modify their views about what form of genetic hypothesis are allowable.”

Do the new concepts apply only to this particular case?

• Does this view of scientific practice applies only for this specific instance of theories we’ve reviewed?• Kitcher thinks that it is generally applicable, more than the concept

of theory as a set of statements:• “I believe that the notion of scientific practice… will both prove helpful in

analyzing the structure of science and the growth of scientific knowledge even in those areas of science where traditional views have seemed most successful” (p. 368)• “I hope that it introduces concepts of general significance in the project of

understanding the growth of science” (p. 369)

Finding the connection between the 2 genetics

• So far we’ve shown that:• there is no reduction from classical to molecular genetics• classical genetics is a theory that’s defined as a scientific practice and without laws

• Powered with the above information, Kitcher wants to find the relation between the two theories. (the original question)• Kitcher moves to list 3 of the major achievements of molecular genetics:

• Replication• Characterization of mutation• Action of particular genes

• Will show for each achievement it’s actual connection to classical genetics. A connection that is not a reduction.• “The three examples reflect three different relations among successive theories, all of

which are different from the classical notion of reduction” (p. 361)

Molecular genetics - replication• Classical genetics believes that all genes replicate, but has no account of gene

replication. • “The claim that genes can replicate does not have the statue of law in classical genetics”• “Instead it is a claim classical geneticists took for granted, a claim presupposed by

explanation – thus we are dealing with presuppositions instead of laws” (p. 361)• Presupposition in a practice is such if there is a problem-solving pattern of the theory that

implies the presupposition (ibid)• “For any problem-solution offered by any version of the theory of gene transmission…will contain

sentences implying that the alleles which it discusses are able to replicate.”

Molecular genetics – replication (cont.)

• Replication is explainable in molecular genetics - “One can show that genes can replicate by showing that any segment of DNA can replicate”(p. 363). • “Contrary to reduction, we do not derive the presuppositions as conclusions in

molecular genetics” (ibid) – there is no reduction here.• “Where the reductionist identifies a general benefit in deriving all the axioms of

the reduced theory, I focus on particular derivation of a claim that no title as an axiom of classical genetics.” (ibid)• “The reductionist’s global relation between theories does not obtain between

classical and molecular genetics, but something akin to it does hold between special fragments of these theories” (ibid)• Molecular genetics provide justification to the presupposition

Molecular genetics - mutation• In classical genetics, “mutant allele was initially fixed through the description

‘chromosomal segment producing a heritable deviant phenotype’” (p. 364)• Provides no account for what mutant alleles are.

• In molecular genetics, “ When we understand the gene as a segment of DNA we recognize the ways in which mutant allele scan be produced” (p.359)• Mutant alleles can be produced by copying errors in replication• “Molecular genetics offers a precise account of the internal changes… with the description more

informative” (p. 364)• Molecular genetics can be seen here to have done a conceptual refinement• “Later theories can be said to provide conceptual refinements when the later theory yields a

specification of entities that belong to the extensions of predicates in the language of the earlier theory, with the result that that the ways in which the referent of these predicates are fixed are altered in accordance with the new specification” (p. 364)

Molecular genetics - action of particular genes

• If we know the molecular characterization of alleles, can we say what the phenotypes would be? • Possible plan for reduction: “We might hope to discover a pattern of reasoning within

molecular genetics that would generate as its conclusion the schema for assigning phenotype to genotypes” (p. 365)• The above plan for reduction fails:

• Molecular genetics are an explanatory extension of classical genetics – “a theory T’ provides explanatory extension of a theory T just in case there is some problem-solving pattern of T whose schematic premises can be generated as the conclusion of a problem-solving pattern of T’” (ibid)

• “However, it does not follow that the explanations provided by the old theory can be improved by replacing the premises in question with the pertinent derivations” (ibid)

• “The explanatory extension does not require any general characterization of genes in molecular terms. All that is needed is the possibility of deriving phenotypic descriptions from molecular characterizations of structures of particular genes” (p. 366)

Molecular genetics - action of particular genes

• Molecular genetics does not always provide an explanatory extension• “Nevertheless, even born-again reductionism is doomed to fall short of salvation.

Although it is true that molecular genetics belongs to a cluster of theories which, taken together, provide an explanatory extension of classical genetics, molecular genetics, on its own, cannot deliver the goods. There are some cases in which the ancillary theories do not contribute to the explanation of a classical claim about gene action. In such cases, the classical claim can be derived and explained by instantiating a pattern drawn from molecular genetics. The example of human hemoglobin provides one such case. But this example is atypical” (p. 366)• In addition, molecular genetics is not always enough to be the explanatory

extension, and additional theories are needed

The relation• We’ve seen that with molecular genetics we can:• Derive presuppositions in classical genetics• Refine concepts in classical genetics• Give an explanatory extension to classical genetics

Prominent views about reduction in biology

• Reductionism vs. Anti-reductionism• Common ground:• “Both reductionism and anti-reductionism in a certain minimal physicalism” (p. 369)• Acknowledgement of levels in nature - both agree on that:• “In evolutionary theory, the anti-reductionists contend that is it impossible to regard all

selection as operating at the level of the gene” (p.369)• “A sophisticated reductionist ought to allow that, in the current practice of biology,

nature is divided into levels..: molecular biology, cytology, histology…each can be thought of as using certain language to formulate the questions it deems important and supplying patterns of reasoning for resolving those questions.” (p. 370)

Reductionism2 possible views:

• Stronger thesis – “the explanation provided by any biological theories can be reformulated in the language of molecular biology and be recast so as to instantiate the patterns of reasoning supplied by molecular biology” (p. 370)• Weaker thesis – “molecular biology provides explanatory extension of

the other biological sciences” (ibid)• Implies a unidirectional flow of explanation• As seen, we can derive presuppositions, and refine concepts

Anti-reductionism2 possible views:• Weaker view: “Anti-reductionism emerges as the thesis that there are

autonomous levels of biological explanation” (p. 371)• This refutes reduction

• “…They can also resist the weaker reductionist view that explanation always flows from the molecular level up” (ibid)• “anti-reductionists can resist the picture of a unidirectional flow of explanation… requires

shifting back and forth across levels.” (ibid)• Gives a case from the world of embryology (p.372)• “We find examples in which claims at a more fundamental level (claims about gene

expression) are to be explained in terms of claims at a less fundamental level (relative positions of pertinent cells)” (ibid.)

Anti-reductionism (cont.)

Corollary: “the explanation provided by the “less fundamental” biological sciences are not extended by molecular biology alone.” (p. 373)

Questions• Vague terms that that, while useful to avoid reduction, seem to create

some sort of an alternative relation “akin to” reduction:• Explanatory extension• Conceptual refinement • Presupposition vs. law• Scientific practice vs. theory• Explanatory levels

• What’s the relation between mechanisms and the intertheory reduction? • What will happen if we discover emergent properties? How does that fit an

anti-reductionist explanation such as Kitcher’s?• Does reductive explanation work only for organized systems (e.g. Level of

cell) or also for aggregative systems (e.g. population of organisms)

Muller’s hope revisited• Has it been fulfilled? • Kitcher believe it has:• “Despite the immense value of the molecular biology… (it) cannot cannibalize

the rest of biology” (p. 373)• “Even if geneticists must become “physiological chemists” they should not

give up being embryologists, physiologists and cytologists”

Bibliography• Nagel, E. (1961). “The Structure of Science”• Kitcher, P. (1984). “1953 and all that. A tale of two sciences”• Godfrey-Smith, P. (2016). “Philosophy of Biology”

Thank you!