Anda di halaman 1dari 8

The British Society for the Philosophy of Science

Putnam on Incommensurability
Author(s): Paul Feyerabend
Reviewed work(s):
Source: The British Journal for the Philosophy of Science, Vol. 38, No. 1 (Mar., 1987), pp. 75-
81
Published by: Oxford University Press on behalf of The British Society for the Philosophy of Science
Stable URL: http://www.jstor.org/stable/687081 .
Accessed: 20/09/2012 09:29

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Oxford University Press and The British Society for the Philosophy of Science are collaborating with JSTOR to
digitize, preserve and extend access to The British Journal for the Philosophy of Science.

http://www.jstor.org
Brit.J. Phil. Sci. 38 (1987),75-92 Printedin GreatBritain 75

Discussions
A RETRACTION OF 'A GEDANKEN EXPERIMENT TO MEASURE THE
ONE-WAY VELOCITY OF LIGHT'

My (Nissim-Sabat, 1984) paper proposing a method for measuring the


speed of light on a one-way trip is wrong and I am retracting it. Briefly, I
had proposed that if one had two identical clocks at A and B respectively
one could measure the speed of light from A to B: the fact that the two
clocks could not be synchronized would be remedied by the interchange of
the positions of the two clocks. Winnie [1970] has shown that if the two
clocks are transported in exactly the same way (i.e., with identical velocities
as measured by round trip methods), then the two clocks would show
exactly the same time lapse (including relativistic time dilation) for their
respective trips. Thus I had argued that any inherent time difference
between the two clocks having been preserved, its effect would have been
cancelled out after the interchange. Yet, following Winnie's reasoning, one
can show that all observations made this way would lead experimenters to
conclude that the velocity of light from A to B is the same as that from B
to A, regardless of whether this is indeed the case. One objection that one
can make to my method is to point out that I had assumed that laboratory
bound observers would necessarily agree with the moving clocks that the
trip from A to B took the same time as that from B to A.
I was led to the above reconsiderations by the objections of an anonymous
referee for the American fournal of Physics to another synchronization
method I had presented.

CHARLES NISSIM-SABAT
Northeastern Illinois University

REFERENCES

NISSIM-SABAT, C. [1984]: 'A Gedanken Experiment to Measure the One-Way Velocity of


Light'. Brit. J. Phil. Sci., 35, 62-64.
WINNIE,J. [1970]: 'Special Relativity Without One-Way Assumptions'. Phil. Sci., 37, 81-99
and 223-238.

PUTNAM ON INCOMMENSURABILITY

(I) In Putnam [1981], p. I 14 Putnam asserts that 'both of the two most
influential philosophies of science of the Twentieth Century . . . are self
76 P. Feyerabend

refuting'. The philosophies he has in mind are logical positivism and the
historical approach. I shall discuss an idea that belongs to the latter viz.
incommensurability and I shall show that while the idea may have unusual
consequences, self refutation is not one of them.
(2) According to Putnam 'the incommensurability thesis is the thesis
that terms used in another culture, say, the term "temperature" as used by
a seventeenth century scientist, cannot be equated in meaning and reference
with any terms or expressions we possess' ( 14). I shall call the incom-
mensurability thesis as defined in this statement T.
To refute T Putnam points out,

(A), that 'if [T] were really true, then we could not translate other languages-or
even past stages of our language-at all' (I 14), adds,
(B), that 'if Feyerabend . . . were right, then members of other cultures, including
seventeenth century scientists would be conceptualisable by us only as animals
producing responses to stimuli' and concludes,
(C), 'To tell us that Galileo has "incommensurable" notions and then to go on and
to describethem at length is totally incoherent' (I 14f-Putnam's italics).

(3) A, B and C rest on the following two assumptions: understanding


foreign concepts (foreign cultures)
[i] requires translation and
[ii] a successful translation does not change the translating language.
Neither [i] nor [ii] is correct. We can learn a language or a culture from
scratch, as a child learns them, without detour through our native tongue
(linguists, historians and anthropologists, having realised the advantages of
such a procedure now prefer field studies to the reports of bilingual infor-
mants). And we can change our native tongue so that it becomes capable
of expressing alien notions (successful translations always change the
medium in which they occur: the only languages satisfying [ii] are formal
languages and the languages of tourists).
Modern lexica exploit both possibilities. Instead of the semantic equa-
tions that formed the basis of older dictionaries they employ research
articles of an open and speculative nature. (An example are the introduction
and the major research articles of Snell [1971].) Analogies, metaphors,
negative characterizations, bits and pieces of cultural history are used to
present a new semantic landscape with new concepts and new connections
between them. Historians of science proceed in a similar way, but more
systematically. Explaining, say, the notion of 'impetus' in I6th and 17th
century science they first teach their readers the physics, metaphysics,
technology, even the theology of the time, i.e. they again introduce a new
and initially unfamiliar semantic landscape and then show where impetus
is located in it. Examples are found in the work of Pierre Duhem, Anneliese
Maier, Marshall Clagett, Hans Blumenberg and, for other concepts, the
work of Ludwik Fleck and Thomas Kuhn.
Translating a language into another language is in many ways like con-
Putnam on Incommensurability 77
structing a scientific theory; in both cases we must find concepts that fit the
'language of the phenomena'. In the natural sciences the phenomena are
those of inanimate nature. Nobody doubts that it is difficult to give a general
account of them, that we may have to revise the terms with which we started
and that we may have to revise them further when new phenomena appear.
In the case of translation the phenomena are the ideas implicit in another
language. These ideas developed in different and often unknown geo-
graphical surroundings, under different and often unknown social cir-
cumstances and they went through numerous intended and unintended
changes (influence of further languages, deterioration, poetic licence, etc.).
Putnam's [ii] assumes that every language contains everything that is
needed for dealing with all these eventualities. To use an example, it makes
the rather unlikely assumption that modern Suaheli is already adapted to
the language of the Eskimo and, therefore, to Eskimo history. There are
only two ways in which such an assumption could succeed: apriorism, or
preestablished harmony. Being an empiricist I reject both.
(4) According to Putnam T makes it impossible to explain foreign (primi-
tive, technical, ancient) concepts in English-this is the content of C. He
is right in one sense, wrong in another. It is indeed impossible, and trivially
so, to formulate ideas in a language not fit to receive them. But the criteria
which identify a natural language do not exclude change. English does not
cease to be English when new words are introduced or old words given a
new sense. Every philologist, anthropologist, sociologist who presents an
archaic (primitive, exotic, etc.) world view, every popular science writer
who wants to explain unusual scientific ideas in ordinary English, every
surrealist, dadaist, teller of fairy tales, ghost stories, science fiction novels,
every translator of the poetry of different ages and nations knows how first
to construct, out of English words, an English sounding model of the pattern
of usage he needs and then to adopt the pattern and to 'speak' it. A rather
trivial example is Evans-Pritchard's explanation of the Azande word mbi-
simo designating the ability of their poison oracle to see far-off things. In
his [1975], p. 151 Evans-Pritchard 'translates' mbisimo as 'soul'. He adds
that it is not soul in our sense, implying life and consciousness but a
collection of public or 'objective' events. The addition modifies the use of
the word 'soul' and makes it more suitable for expressing what the Azande
had in mind. Why 'soul' and not another word? 'Because the notion this
word expresses in our own culture is nearer to the Zande notion of mbisimo
of persons than any other English word'-i.e. because of an analogy between
the English soul and the Azande mbisimo. The analogy is important for it
smoothes the transition from the original to the new sense; we feel that
despite the change of meaning we are still speaking the same language. Now
if a conceptual change like the one just described does not go through a
metalanguage but stays in the language itself (in which case we would speak
of changing the properties of things rather than the usage of words) and if
it is not only a single term but an entire conceptual system that is being
78 P. Feyerabend
received then we have the situation alluded to in (C), but defused, for the
English with which we start is not the English with which we conclude our
explanation.
(5) Azande ideas already exist in a spoken language and English notions
were changed to accommodate them. There are other cases where linguistic
change introduces a novel and as yet unexpressed point of view. The history
of science contains many examples of this kind. I shall explain the matter
by taking an example from the history of ideas.
In Iliad 9, 225ff Odysseus tries to get Achilles back into the battle against
the Trojans. Achilles resists. 'Equal fate' he replies 'befalls the negligent
and the valiant fighter; equal honor goes to the worthless and to the vir-
tuous' (3 I8f). He seems to say that honor and the appearance of honor are
two different things.
The archaic notion of honor did not allow for such a distinction. Honor,
as understood in the epic was an aggregate consisting of partly individual
partly collective actions and events. Some of the elements of the aggregate
were: the position (of the individual possessing or lacking honor) in battle,
at the assembly, during internal dissension; his place at public ceremonies;
the spoils and gifts he received when a battle was finished and, naturally,
his behavior on all these occasions. Honor was present when (most of) the
elements of the aggregate were present, absent otherwise (cf. II. 12, 3 Ioff--
Sarpedon's speech).
Achilles introduces a different point of view. He was offended by Aga-
memnon who took his gifts. The offence created a conflict between the
individual and the collective ingredients of honor. The Greeks who appeal
to Achilles, Odysseus among them, illustrate the customary resolution of the
conflict: Achilles' gifts have been returned, more gifts have been promised,
harmony has returned to the aggregate, honor has been restored (519, 526,
602f). So far we are squarely within tradition. Achilles moves away from it.
Pushed along by his lasting anger he perceives an equally lasting imbalance
between personal worth and social rewards. What he has in mind not only
differs from the traditional aggregate, it is not even an aggregate for there
is no set of events that guarantees the presence of honor as he now sees it.
Using Putnam's terminology we can say that Achilles' idea of honor is
'incommensurable' with the traditional idea. And indeed, given the epic
background the short excerpt I quoted from Achilles' speech sounds as
nonsensical as the statement 'equal time needs the fast and the slow to reach
the goal'. Yet Achilles introduces his idea in the very same language that
seems to exclude it. How is that possible?
It is possible because, like Evans-Pritchard, Achilles can change concepts
while retaining the associated words. And he can change concepts without
ceasing to speak Greek because concepts are ambiguous, elastic, capable
of reinterpretation, extrapolation, restriction, or, to use a term from the
psychology of perception, concepts like percepts obey figure-ground
relations.
Putnam on Incommensurability 79
For example, the tension between the individual and the collective
elements of honor that was caused by Agamemnon's deed can be seen in at
least two ways, as involving ingredients of equal weight, or as a conflict
between fundamental and more peripheral elements. Tradition accepted
the first view or, rather, there was no question of a conscious acceptance-
people simply acted that way: 'With gifts promised go forth. The Achaians
will honor you as they would an immortal!' (602f). Achilles driven by his
anger magnifies the tension so that it turns from a transitory disturbance
into a cosmic rift (figure-ground relations often change as a result of strong
emotions; this is the principle of the Rohrschach-test). The extrapolation
does not void his speech of meaning because there exist analogies for what
he is trying to express. Divine knowledge and human knowledge, divine
power and human power, human intention and human speech (an example
used by Achilles himself: 312f) are opposed to each other as Achilles
opposes personal honor and its collective manifestations. Guided by the
analogies his audience is drawn into the second way of seeing the tension
and so discovers, as Achilles did, a new side of honor and of archaic morality.
The new side is not as well defined as the archaic notion-it is more a
foreboding than a concept-but the foreboding produces new ways of
speaking and thus, eventually, clear new concepts (the concepts of some
Presocratic philosophers are endpoints of this line of development). Taking
the unchanged traditional concepts as a measure of sense we are of course
forced to say that Achilles speaks nonsense (cf. Parry [1956] for this point;
cf. also Feyerabend [1975], 267). But measures of sense are not rigid and
unambiguous and their changes are not so unfamiliar as to prevent the
listeners from grasping what Achilles has in mind. Speaking a language or
explaining a situation, after all, means both following rules and changing
them; it is an almost inextricable web of logical and rhetorical moves.
From what has just been said it also follows that speaking a language
goes through stages where speaking indeed amounts to merely 'making
noises' (Putnam [I98i], 122). For Putnam this is a criticism of the views
he ascribes to Kuhn and myself. For me it is a sign that Putnam is unaware
of the many ways in which language can be used. Little children learn a
language by attending to noises which, being repeated in suitable sur-
roundings, gradually assume meaning. Commenting on the explanations
which his father gave him on logical matters Mill writes (Lerner [19651,
21): 'The explanations did not make the matter clear to me at the time;
but they were not therefore useless; they remained as a nucleus for my
observations and reflections to crystallize upon; the import of his general
remarks being interpreted to me, by the particular instances which came
under my notice afterwards.' Saint Augustine advised parsons to teach the
formulae of the faith by rote adding that their sense would emerge as a
result of prolonged use within a rich, eventful and pious life. Theoretical
physicists often play around with formulae which do not yet make any sense
to them until a lucky combination makes everything fall into place (in the
80 P. Feyerabend
case of the quantum theory we are still waiting for this lucky combination).
And Achilles, by his way of talking, created new speech habits which
eventually gave rise to new and more abstract conceptions of honor, virtue
and being. Thus using words as mere noises has an important function
even within the most advanced stages of speaking a language (cf. Fey-
erabend [ 19751, 270).
One of the scientists who was aware of the complex nature of explanatory
talk and who used its elements with superb skill was Galileo. Like Achilles
Galileo gave new meanings to old and familiar words; like Achilles he
presented his results as parts of a framework that was shared and understood
by all (I am now speaking of his change of basic kinematic and dynamical
notions); but unlike Achilles he knew what he was doing and he tried to
conceal the conceptual changes he needed to guarantee the validity of his
arguments. Chapters 6 and 7 of my [1975] contain examples of his art.
Taken together with what I have said up till now these examples show how
it is possible to assert, without becoming incoherent, that the Galilean
notions are 'incommensurable' with our own 'and then to go on and to
describe them at length'.
(6) They also solve Putnam's conundrum about the relation between
relativity and classical mechanics. If T is correct, says Putnam, then the
sense of statements that occur in a test of either relativity or classical
mechanics cannot be 'independent of the choice between Newtonian and
Einsteinian theory'. Moreover, it is then impossible to find equations of
meaning between 'any word in ... Newtonian theory [and] any word in ...
general relativi[ty]' (i 16). He infers that there is no way to compare the two
theories. The inference is again mistaken. As I mentioned in section 3
linguists long ago ceased using equations of meaning to explain new and
unfamiliar ideas while scientists always emphasized the novelty of their
discoveries and of the concepts used in their formulation. This does not
stop them from comparing theories, however. Thus the relativist can say
that the classical formulae, properly interpreted (i.e. interpreted in the rela-
tivistic manner) are successful, but not as successful as the full relativistic
apparatus. He can argue like a psychiatrist who, talking to a patient who
believes in demons (Newton) adopts his, the patient's manner of speaking
without accepting its demonic (Newtonian) implications (this does not
exclude the the possibility that the patient will one fine day turn around
and convince him of the existence of demons). Or he may teach relativity
to the classicist like a foreign language and invite him to judge its virtues
from within ('having learned Spanish to perfection and having read Borges
and Vargas Llosa, would you not rather write stories in Spanish than in
German?') There exist many other ways in which the Newtonian and the
relativist can and do converse. I have explained them in papers I wrote in
1965, some of them in direct response to Putnam's criticisms of that time:
Feyerabend [198I], Vol. i, chapter 6, sections 5ff and Vol. ii, chapter 8,
section 9ff and appendix. This finishes my response to A, B and C.
Putnam on Incommensurability 81
(7) The arguments of the preceding sections were based on T which is
Putnam's version of incommensurability. But Putnam's version is not the
version I introduced when examining the relation between comprehensive
theories such as Newton's mechanics and relativity or Aristotelian Physics
and the new mechanics of Galileo and Newton (cf. Feyerabend [1975] 268ff
and [1981], Vol. i, chapter 4, section 5). There are two differences. First,
incommensurability as understood by me is a rare event. It occurs only
when the conditions of meaningfulness for the descriptive terms of one
language (theory, point of view) do not permit the use of the descriptive
terms of another language (theory, point of view); mere difference of mean-
ings does not yet lead to incommensurability in my sense. Secondly, incom-
mensurable languages (theories, points of view) are not completely dis-
connected-there exists a subtle and interesting relation between their
conditions of meaningfulness. In my [1975] I explained this relation in the
case of Homeric Commonsense vs. the language aimed at by the early Greek
philosophers. In [i98I] Vol. i, chapter 4 I explained it in the case of
Aristotle-Newton. I should add that incommensurability is a difficulty for
philosophers, not for scientists. Philosophers insist on stability of meaning
throughout an argument while scientists, being aware that 'speaking a
language or explaining a situation means both following rules and changing
them' (see section 5 of this paper) are experts in the art of arguing across
lines some philosophers regard as insuperable boundaries of discourse.

PAUL FEYERABEND
University of California
at Berkeley

REFERENCES

EVANS-PRITCHARD, E. E. [1976]: Witchcraft, Oracles and Magic Among the Azande, abridged
edition, Oxford, 1976.
FEYERABEND, PAULK. [1975]: Against Method, London, 1975.
FEYERABEND, PAUL K. [1981]: Philosophical Papers, 2 Vols., Cambridge, 1981.
LERNER,MAX [1965]: Essential Works of John Stuart Mill, New York, 1965.
PARRY,A. [1956]: 'The Language of Achilles', Trans. & Proc. Amer. Philos. Assoc., Vol. 87,
1956.
PUTNAM,HILARY[1981]: Reason, Truth and History, Cambridge, 1981.
SNELL, BRUNO[1971]: Lexikon des FruehgrechischenEpos, Goettingen, 1971.

PAPINEAU ON CAUSAL ASYMMETRY

David Papineau in 'Causal Asymmetry' attempts to provide an account of


causal asymmetry which does not depend upon the assumption that causes
precede their effects in time [1985]. In place of a temporal-priority view,
he introduces a probabilistic analysis. This analysis takes two forms, one