Laura FranklinHall New York University April 2015 Scientific Classifications g ranular materials f undamental particles g enes e ventoed ungulates NigerCongo languages Differences Between Classifications ID: 418305
Download Presentation The PPT/PDF document "Why are some kinds Historical and Others..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Why are some kinds Historical and Others Not?
Laura Franklin-Hall
New York University
April 2015Slide2
Scientific Classifications
g
ranular materials
f
undamental particles
g
enes
e
ven-toed ungulates
Niger-Congo languagesSlide3
Differences Between Classifications
Fine-grained or coarse-grained
Structural vs. functional
Architectonic vs. simple
Essentialist vs. non-essentialist
Historical vs. synchronicSlide4
Fine-grained or coarse-grained
Structural vs. functional
Architectonic vs. simple
Essentialist vs. non-essentialist
Historical vs. synchronic
Differences Between Classifications
A
full theory of classification
would account for all of these differencesSlide5
Fine-grained or coarse-grained
Structural vs. functional
Architectonic vs. simple
Essentialist vs. non-essentialist
Historical vs. synchronic
Differences Between Classifications
My focus
:
to explore and rationalize this differenceSlide6
Plan
Step 1: Characterize synchronic vs. historical
Step 2: Ask why classifications are sometimes historical and sometimes synchronic
Step 3: Use the Probability-Similarity Account (PSA) to answer this question
Step 4: Rationalize and defend the PSA
Step 5: Consider implications for the mind-dependence of the natural kindsSlide7
1: Synchronic vs. Historical
Synchronic Classification
Key property is a feature of the classified individuals grounded synchronically
Historical Classification
Key property is a historical feature of the individuals classified – usually descent via reproduction
Not exhaustive: there could be a ‘future classification’ where things are grouped in terms of what they will bring about.
Also, there are mixed cases: e.g
., metamorphic
rocksSlide8
1: Synchronic vs. Historical
Synchronic Classification
Historical Classification
Bacterial species by
nuclear DNA sequence
Molecules by
constituent atoms plus bonding
Gaits by
leg-lift patterns
Metazoan species by descent from a founding population
Words by sound / meaning hybrid in a language, with languages individuated by history
Cultural items by
originsSlide9
1: Synchronic vs. Historical
Synchronic Classification
Historical Classification
Bacterial species by
nuclear DNA sequence
Molecules by
constituent atoms plus bonding
Gaits by
leg-lift patterns
Metazoan species by
descent from a founding population
Words by
sound / meaning hybrid in a language, with languages individuated by history
Cultural items by
origins
To be explored later….Slide10
1: Synchronic vs. Historical
Synchronic Classification
Bacterial species by
nuclear DNA sequence
Molecules by
constituent atoms plus bonding
Gaits by
leg-lift patterns
Caffeine
Formula: C
8
H
10
N
4
O
2
IUPAC
ID: 1,3,7-Trimethylpurine-2,6-dione
To count as caffeine, a substance must just be made of such molecules. It’s origin can be either natural or artificial.Slide11
1: Synchronic vs. Historical
Synchronic Classification
Bacterial species by
nuclear DNA sequence
Molecules by
constituent atoms plus topology
Gaits by
leg-lift patterns
The gallop
A four-beat gait in which all feet are briefly off the ground at the same time
For a motion to count as a gallop, it doesn’t matter how the animal learned the process or anything else about its ontogeny.Slide12
1: Synchronic vs. Historical
Historical Classification
Metazoan species by
descent from a founding population
Words by
meaning/ sound hybrid in a language, with languages individuated by history
Cultural items by
origins
Word
Language Tree
+
in a language (defined by origins)
‘False cognates are not instances of the ‘same word’
i.e., ‘much’ and ‘mucho’Slide13
1: Synchronic vs. Historical
Historical Classification
Metazoan species by
descent from a founding population
Words by
meaning/ sound hybrid in a language, with languages individuated by history
Cultural items by
origins
Holiday – The Day of the Dead (
Dia
de Muertos)
Originated in southern Mexico, from a Aztec rite.
Hybridized with All Souls Day.
Spread through
L
atin
A
merica where the ‘same holiday’ is celebrated.
Similar in practice and belief to
other
holidays
found elsewhere (e.g., the Ghost Festival in China); also involve honoring ancestors. Slide14
1: Synchronic vs. Historical
Caveats
controversy about the ‘key property’ for some classifications. I have tried to presume the dominant view.
In biology, controversy about whether classifications correspond to kinds or individuals. I don’t think this difference makes a difference and will put it aside.Slide15
2: the explanatory Question
Why are some categories
in use by scientists
historical and others synchronic?
(I say ‘in use by’ because there is some sense that both historical and synchronic categories could be applied
to any science. The question is which kind of category actually gets developed.)
Presuming, as many naturalistic philosophers do, that the natural kinds correspond to the categories in use by mature sciences, this question also permits us to query why some natural kinds are historical and others not
.Slide16
2: the explanatory Question
Why are some categories
in use by scientists
historical and others synchronic?
Obvious reply
“Categories are historical in historical sciences and synchronic otherwise”
No: all historical sciences that aim to account for causal change will require synchronic categories as well
Biology, Linguistics and Anthropology all have synchronic categories alongside their historical ones.Slide17
3: The Probability-Similarity Account
My proposal: for any domain of individuals, scientific classifications of those individuals will be historical whenever
The
P
robability
of the Independent
Emergence of
Similar things (PIES) is very small
They are synchronic whenever PIES is not very smallSlide18
3: The Probability-Similarity Account
The
Probability
of the
Independent
Emergence of Similar things (
PIES)Slide19
3: The Probability-Similarity Account
The
Probability
of the
Independent
Emergence of Similar things (
PIES)
What does this mean?Slide20
3: The Probability-Similarity Account
What does this mean?
The
Probability
of
the Independent Emergence
of Similar things (PIES)Slide21
3: The Probability-Similarity Account
time
What does this mean?
The
Probability
of
the
Independent
Emergence
of
Similar
things (
PIES)Slide22
3: The Probability-Similarity Account
time
What does this mean?
The
Probability
of
the
Independent
Emergence of Similar things (PIES)Slide23
3: The Probability-Similarity Account
time
What does this mean?
single event
Case 1:
not independent
Most likely: reproductive lineage
The
Probability
of
the
Independent
Emergence
of
Similar
things (
PIES)Slide24
3: The Probability-Similarity Account
time
What does this mean?
e
vent 2
Case 2: independent
e
vent 1
The
Probability
of
the
Independent
Emergence
of
Similar
things (
PIES)Slide25
3: The Probability-Similarity Account
time
What does this mean?
e
vent 2
Case 2: independent
e
vent 1
N.B.: These cartoons paper over many issues, some of which will be addressed shortly…
The
Probability
of
the
Independent
Emergence
of
Similar
things (
PIES)Slide26
3: The Probability-Similarity Account
time
What does this mean?
e
vent 2
Case 2: independent
e
vent 1
The
Probability
of the
Independent Emergence of
Similar
things (
PIES)
PIES is high when CASE 2 is likely
Synchronic kinds
PIES is low when CASE 2 is unlikely
Historical Kinds
We often judge probabilities based on frequencies, but they can come apart.Slide27
3: The Probability-Similarity Account
Illustration: atoms
Probability that similar individual atoms (e.g., elements) comes about in independent origination events is
high
. Thus, their kinds (
the elements) are synchronic.Evidence: frequencies track probabilities and many atoms of any elements have had separate originations; some in one supernova, others in another, others via radioactive decay. Slide28
3: The Probability-Similarity Account
Illustration: atoms
Probability that similar individual atoms (e.g., elements) comes about in independent origination events is
high
. Thus, their kinds (
the elements) are synchronic.
Explanation: unlike with solar systems: small variations in initial conditions for the generation of an atom can still yield identical atoms; this traces ultimately to the quantum nature of our universe (see Ghirardi
2007) Slide29
3: The Probability-Similarity Account
Illustration: organisms
Probability that similar individual organisms (e.g., species) comes about independently is
low
. Thus, organisms are divided into
historical kinds: the species.Evidence: frequencies track probabilities and famously there are no instances of strong convergence between organisms with different origins.Slide30
3: The Probability-Similarity Account
Illustration: organisms
Probability that similar individual organisms (e.g., species) comes about independently is
low
. Thus, organisms are divided into
historical kinds: the species.Explanation: Many small and undirected modifications are required to result in any given form. That these would be identical modifications in two lines is highly unlikely.Slide31
3: The Probability-Similarity Account
The
Probability
of the
Independent
Emergence of Similar things (
PIES)Slide32
3: The Probability-Similarity Account
The
Probability
of the
Independent
Emergence of Similar things (
PIES)
REACTION:
SOUNDS PLAUSIBLE, BUT WHAT DETERMINES THESE PROBABILIIES?Slide33
3: The Probability-Similarity Account
The
Probability
of the
Independent
Emergence of Similar things (
PIES)
MY ANSWER:
A COMBINATION OF THE CAUSAL FEATURES OF SYSTEMS AND
FEATURES OF US: OUR ‘GAZE’, OUR NOTION OF INDEPENCENCE,
OUR STANDARDS FOR SIMILARITIES
REACTION:
SOUNDS PLAUSIBLE, BUT WHAT DETERMINES THESE PROBABILIIES?Slide34
4: Defending the Account
Two-part defense of the PIES standard:
Saves the phenomena: looks at different scientific examples, and see whether PIES tracks classification-type
Rationalize the principle: argue that it makes sense for practice to respect itSlide35
4: Defending the Account
Two-part defense of the PIES standard:
Saves the phenomena: looks at different scientific examples, and see whether PIES tracks classification-typeSlide36
4: Defending the Account
The case of biology
1. Eukaryotic Species
2. Bacterial Species
3. Plant Species
4
. Kinds in Astrobiology
5. Kinds for Convergence-Lovers6. Lamarckian Species
Saving the PhenomenaSlide37
4: Defending the Account
1.
Eukaryotic Species
Saving the Phenomena
The species of an organism is most often taken to be
a function of its history
(phylogenetic species concept)
An organism’s synchronic properties are not relevant to questions of its species membership
CLASSIFICATION TYPE: HISTORICALSlide38
4
: Defending the Account
1.
Eukaryotic Species
Saving the Phenomena
‘swamp pig’
not
a pig
‘’snail’-born-of-pig’
is
a pig
Mom!
CLASSIFICATION TYPE: HISTORICAL
The species of an organism is most often taken to be
a function of its history
(phylogenetic species concept)
An organism’s synchronic properties are not relevant to questions of its species membershipSlide39
4: Defending the Account
1.
Eukaryotic Species
Saving the Phenomena
PIES Analysis: The probability of the evolution of any very complex package of traits, such as are found among pigs, is extreme low. (
cf
‘
Dollo’s
law’)
historical kinds
EXPLANATIONSlide40
4: Defending the Account
2.
Bacterial Species
Saving the Phenomena
Bacterial ‘tree’
T
he ‘official definition’ of bacterial species depends on similarity between genomes*.
An organism’s history
is not strictly relevant
to species membership.
*70% DNA molecule cross-hybridization, which occurs with approximates 95% nucleotide identity
CLASSIFICATION TYPE: SYNCHRONICSlide41
4
: Defending the Account
2.
Bacterial Species
Saving the Phenomena
Bacterial ‘tree’
swamp E. coli and
E.coli
-borne-of-staph
are
E. coli
CLASSIFICATION TYPE: SYNCHRONIC
Mom!
staph
T
he ‘official definition’ of bacterial species depends on similarity between genomes*.
An organism’s history
is not strictly relevant
to species membership. Slide42
4: Defending the Account
2.
Bacterial Species
Saving the Phenomena
Top: transduction
Bottom: conjugation
Speciation via non-reproductive gene transfer (aka Lateral Gene Transfer or LGT).
This makes it
comparatively
easy to get the same set of traits in separate lineages
EXPLANATIONSlide43
4: Defending the Account
2.
Bacterial Species
Saving the Phenomena
Top: transduction
Bottom: conjugation
PIES Analysis: The probability of the repeated independent emergence of a particular package of traits, such as are found among E. coli, is not extremely low
synchronic kinds
Speciation via non-reproductive gene transfer (aka Lateral Gene Transfer or LGT).
This makes it
comparatively
easy to get the same set of traits in separate lineages
EXPLANATIONSlide44
4: Defending the Account
3
.
Plant Species
Saving the Phenomena
Recurrent speciation in some plant species; botanists studying these species consider products to be
single species with multiple origins
.
“The same species can actually form multiple times”
“Recurrent speciation […] is the rule, not the exception”
CLASSIFICATION TYPE: SYNCHRONICSlide45
4: Defending the Account
3
.
Plant Species
Saving the Phenomena
Hybridization
PIES Analysis: Hybrid
speciation increases likelihood of similar entities with ‘independent’
origins
Synchronic kinds
T.
dubius
, T.
pratensis
, and T.
porrifolius
were introduced from Europe into Eastern Washington/Idaho in the early 1900s. Two
polyploid
hybids
, T.
miscellus
and T.
mirus
, came about many times when ranges overlapped.
EXPLANATIONSlide46
4: Defending the Account
4.
Astrobiological
Groupings (‘species’?)
Saving the Phenomena
Astrobiologists
—unlike most others who think about the life sciences—sometimes talk of a ‘periodic table of organisms’.
They don’t know just what the bins in this table will be due to the n=1 problem (i.e., our only examples of living things are from earth)
But they imagine that being in a bin is grounded synchronically, not via history.
CLASSIFICATION TYPE: SYNCHRONICSlide47
4: Defending the Account
4.
Astrobiological
Groupings (‘species’?)
Saving the Phenomena
EXPLANATION
Focus on such an enormous sample space, in concert with low standards of similarity
PIES Analysis: Broad sample space makes similar organisms with independent origins less unlikely
s
ynchronic kindsSlide48
4: Defending the Account
5
.
Categories among the convergence-obsessed
Saving the Phenomena
H
eterodox biologists (Conway Morris and McGhee) have also tried to design ‘a periodic table of life’.
Membership in a bin depends on synchronic properties, not history.
CLASSIFICATION TYPE: SYNCHRONICSlide49
4: Defending the Account
5
.
Categories among the convergence-obsessed
Saving the Phenomena
EXPLANATION
These scientists believe that strong convergence is likely (e.g., similar forms will evolve via very different lineages). They also have in mind comparatively weak standards for similarity.
PIES Analysis: Peculiar causal theory + weak standard for similarity makes PIES high
s
ynchronic kindsSlide50
4: Defending the Account
6
.
Lamarckian ‘Species’
Saving the PhenomenaSlide51
4: Defending the Account
6
.
Lamarckian ‘Species’
Saving the Phenomena
Chapter 4,
On the Origin of Species
DARWIN’S VIEW:
LIFE’S TREE
timeSlide52
4
: Defending the Account
Saving the Phenomena
time
present
LAMARCK’S VIEW:
LIFE’S REEDS
6
.
Lamarckian ‘Species’
These are separately-originating ‘reed-like’ lineagesSlide53
4
: Defending the Account
Saving the Phenomena
time
Similar individuals can have independent origins. When L talks about ‘species’ he is using a synchronic conception. He has a false view of origins, but he is still following the PIES principle.
present
6
.
Lamarckian ‘Species’
CLASSIFICATION TYPE: SYNCHRONIC
These are separately-originating ‘reed-like’ lineages
LAMARCK’S VIEW:
LIFE’S REEDSSlide54
4: Defending the Account
1. Eukaryotic Species
2. Prokaryotic Species
3. Plant Species
4
. Kinds in Astrobiology
5. Kinds for convergence-lovers 6. Lamarckian Species
The case of biology - RECAP
SYNCHRONIC
SYNCHRONIC
SYNCHRONIC
SYNCHRONIC
SYNCHRONIC
HISTORICAL
Saving the PhenomenaSlide55
4
: Defending the Account
Saving the Phenomena
The
Probability
of the
Independent
Emergence
of Similar
things (PIES)
MY ANSWER:
A COMBINATION OF THE CAUSAL FEATURES OF SYSTEMS AND FEATURES OF US:
OUR ‘GAZE’,
OUR NOTION OF INDEPENCENCE,
OUR STANDARDS FOR SIMILARITIES
REACTION:
SOUNDS PLAUSIBLE, BUT WHAT DETERMINES THESE PROBABILIIES?Slide56
4: Defending the Account
Saving the Phenomena
CAUSAL FACTS (C)
GAZE (G)
SIMILARITY STANDARDS
(S)
INDEPENDENCE CONDITIONS (I)Slide57
4: Defending the Account
The case of biology - RECAP
1. Eukaryotic Species none
2. Prokaryotic Species
causal
dif
, demanding notion of I
3. Plant Species causal dif
, demanding notion of I4. Kinds in Astrobiology
weak S, wide gaze5. Kinds for convergence-lovers causal
dif, weak S6. Lamarckian Species causal
dif
Saving the Phenomena
DIAGNOSISSlide58
4: Defending the Account
The case of linguistics
1. Most words
2. Sounds
3. ‘Huh?’
Saving the PhenomenaSlide59
4: Defending the Account
1. Most words
Saving the Phenomena
Groupings of utterances into words has historical component
Illustration: ‘false cognates’ are not considered to be instances of the same word, even though they can (in the extreme) be signs that both share sound and meaning.
Explanation: arbitrariness of connection between sign and signifier makes sound-meaning combinations with independent origins unlikely
much (13th c.) ←< southern Middle English
muchel
(where “u” is actually front
[y]; 12th c.) < Old English
miċel ‘big’ < Proto-Germanic *mikilaz (cf.Gothic
mikils) < Proto-Indo-European *meǵ- (cf. Greek mégas and Latinmagnus
, both ‘big’, and Hittite
mēk
‘much’)12
Spanish mucho < *
muito
(cf. Portuguese
muito
and the Spanish adverb
muy
,
which was truncated in unstressed position in rapid speech) < Latin
multum
‘much’ < *
mol
-to-; the root is PIE *
mel
-, which appears also in
Latin
melior
‘better’ (> Spanish
mejor
)Slide60
4: Defending the Account
2
. Sounds
Saving the Phenomena
Individual sounds are not classified in a way that depends on history
Most sounds have ‘multiple origins’ in different language systems Slide61
4: Defending the Account
The case of linguistics
3
. ‘Huh?’
Saving the Phenomena
The exception proves the rule: the word ‘huh’ does have independent origins, and is not considered a ‘false cognate’ but instead a ‘universal word’.
Why? Strong ‘selection pressures’ on this utterance made similarity in the face of ‘independent origins’ likely.Slide62
4: Defending the Account
Two-part defense of the PIES standard:
Saves the phenomena: looks at different scientific examples, and see whether PIES tracks classification-type
Rationalize the principle: argue that it makes sense for practice to respect it
STRATEGY: point to two scientific goals accomplished by using historical categories when PIES <<Slide63
4: Defending the Account
Rationalizing the Principle
Use of Categories: Strong Generalizations
e.g., Pandas eat bamboo, electrons have mass of 9.109 x 10
-31
When PIES << the strongest possible generalizations will use historical categories
When PIES ~<< (non-disjunctive) historical categories will not be as strongSlide64
4: Defending the Account
Rationalizing the Principle
Use of Categories:
Explanatoriness
Generalization Strength only explains why historical categories are
coequal
with synchronic ones with PIES << not why they would be
preferred. Explanatoriness
goes further. Historical categories include information that accounts for similarities common to a kind: that is, that these instances have a common origin.
Note: explanatoriness also accounts of which synchronic features are considered ‘key’, as in an explanatory approach to Lockean essences Slide65
Plan
Step 1: Characterize synchronic vs. historical
Step 2: Ask why classifications are sometimes historical and sometimes synchronic
Step 3: Use the Probability-Similarity Account (PSA) to answer this question
Step 4: Rationalize and defend the PSA
Step 5: Consider implications for the mind-dependence of the natural kindsSlide66
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?Slide67
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
Because PIES depends both on the causal systems and features of us (gaze, similarities standards, standard for independence), facts about whether kinds are historical
will not be strongly mind-independent
. This speaks against a certain kind of realism.Slide68
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
Two ways it is consistent with ‘realism’ (properly understood)
Promiscuous realism
Categorical Bottlenecks
Note: focus here on the ‘reality’ of their status as historical or synchronicSlide69
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
Promiscuous realism
Deny the premise, floated above, that grouping corresponding to the classifications of science are the natural kinds.
P
ermit the natural kinds to be promiscuous, to outrun the classifications of science.
Contextual features that determine what kind of classification we out to use then determine simply which kinds to focus on / find relevant, and the subjectivity of PIES will not undermine the kinds mind-independence.Slide70
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
2. Categorical Bottlenecks
See Franklin-Hall, L (2015) “Natural Kinds as Categorical Bottlenecks” Phil StudiesSlide71
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
2. Categorical Bottlenecks
See Franklin-Hall, L (2015) “Natural Kinds as Categorical Bottlenecks” Phil Studies
Epistemic agent space
x
x
x
x
Categorization Space
x
x
x
xSlide72
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
2. Categorical Bottlenecks
See Franklin-Hall, L (2015) “Natural Kinds as Categorical Bottlenecks” Phil Studies
Epistemic agent space
x
x
x
x
x
x
Categorization Space
x
xSlide73
5: Mind-Dependence?
Mind-dependence or ‘reality’ or natural kinds on PIES picture?
2. Categorical Bottlenecks
See Franklin-Hall, L (2015) “Natural Kinds as Categorical Bottlenecks” Phil Studies
Epistemic agent space
x
x
x
x
x
x
Categorization Space
x
x
Eukaryotic species,
e
lements, and molecules have the ‘bottleneck’ structure – more ‘real’
Kinds of plants, bacteria, etc. do notSlide74
Final
(Counter?)
ExamPLE
Higgs Boson
PIES <<
But a synchronic rather than historical kind
Why? Holism in classificatory principlesSlide75
Thanks!