The Voice of America presents Forum the Arts and Sciences in mid-century America today as the 11th in the philosophy of science series Forum brings you Professor Carl G hemple of Princeton University who will speak on scientific explanation born in orian Bor Germany Professor hemple studied mathematics physics and philosophy at the University ities of giringan heidleberg and Vienna and received his Doctorate from the University of Berlin Professor hemple came to the United States in 1937 and has been a member of the faculties of the University of Chicago the College of the city of New York Queens
College and Yale University he joined the faculty of Princeton in 1955 Dr hemple a fellow of the American Academy of Arts and Sciences has has taught at Oxford University as a Fullbright fellow Dr hemple among the many factors that have stimulated and sustained scientific inquiry throughout its long history there are two enduring human concerns that have provided the most important motivation for man's scientific efforts one of these is a practical concern man constantly strives to improve his strategic IC position Visa the world he lives in to this end he seeks to develop ever more effective
ways of predicting the events in the world around him and of controlling their course to his own Advantage how successful scientific research has been on this score is clearly shown by the vast and steadily widening range of its technological applications both construct RVE and destructive which have put their characteristic imprint on every aspect of contemporary civilization the second basic motive for man's scientific Quest is independent of such practical interests it lies in his insatiable intellectual curiosity in his deep desire to know the many facets and changes of the world and to understand why they OCC
her so demanding is this urge for knowledge and understanding that in the absence of adequate factual information myths will often be invoked to answer questions as to the what and the why of the empirical World gradually however such myths are replaced by answers in terms of Concepts and theories that have been arrived at by scientific inquiry scientific inquiry in the wide sense including research in the Natural Sciences as well as in Psychology and in the sociological and historical disciplines what is the nature of the explanations that empirical science in this broad sense can offer us
and what sort of understanding of the world do they convey an explanation in the sense that concerns here is basically an answer to the question why a given event occurred or why a certain State of Affairs obtains questions of this kind are often answered by indicating what caused the given phenomenon but as we have learned from Hume the assertion that an event of a specified kind C causes an event of a certain other kind e as its effect implies is the claim that any occurrence of C is regularly accompanied by An Occurrence of e for
example that in any case when the temperature of a copper Rod is raised its length increases thus causal explanations presuppose general laws of nature that connect the specified cause with the effect to be explained for the explanations just mentioned science can supply the relevant laws these govern the thermal expansion of metals the magnetic effects of electric currents and the relative motion of two bodies under the influence of their Mutual gravitational attraction now if the relevant laws are explicitly stated the resulting explanation can be put into the form of a deductive argument in which the occurrence
of the event in question is inferred from a set of premises which specify on the one hand the relevant laws and on the other those particular antecedent circumstances which in everyday parant are said to have caused the event for example our explanation of the lengthening of a copper Rod would take the form of an argument with two premises first the general law that any copper Rod grows longer when its temperature is raised and second the statement that the given Rod was made of copper and that its temperature had been raised these premises jointly imply deductively
the conclusion that the rod increased in length which is the event to be explained thus the explanatory statement that the lengthening of the rod was caused by the increase in its temperature is replaced here by an argument which no longer contains the word cause or its cognates at all briefly the argument is to the effect that the given copper Rod was heated and all heated copper rods grow longer and that therefore the given Rod increased in length it may generally be said that the technical vocabulary in which scientific hypotheses and theories are expressed does not
contain such words as cause and causal Factor and that an account in terms of causes is acceptable as an explicit explanation in science only as far as the causal attribution it makes can be substantiated and replaced by statements of corresponding laws such laws will normally take a precise quantitative form this is the case for example with the laws for the thermal expansion of metals which accordingly make it possible to explain not only the fact but also the extent of the change in length that occurs in response to temperature changes it should be noted that in
nontechnical everyday language an explanation will often be expressed elliptically by means of a simple because statement which mentions only one or a very few of the many items that would have to be specified if the explanation were to be formulated as an explanatory scientific argument take for example the statement that the moon keeps moving about the Earth because of the mutual gravitational attraction of the two bodies in an explicit restatement of this sketchy account the explanatory premises might include Newton's Laws of gravitation and of motion as as well as particular statements about the masses of
the two bodies and their relative positions and velocities at some particular time and the deduction of the desired conclusion from these premises requires not simple syllogistic reasoning but the powerful mathematical techniques of the calculus explanations of the kind we have just examined may be called explanation by deductive subsumption undercovering laws or briefly deductive explanations a deductive explanation of a given event shows that the event resulted from specified particular circumstances in accordance with certain general laws thus it enables us to understand the event by making us aware that in view of those laws and particular circumstances
it its occurrence was to be expected so far we have considered deductive explanations which in non-technical language would be formulated as causal statements but the explanatory power of deductive subsumption under general laws extends far beyond cases of this kind all our examples so far deal with the explanation of particular events but science seeks to answer the question why not only for this or that individual occurrence but also for the uniformities and regularities expressed by general laws thus in the case of the pendulum we might proceed to ask why does any simple pendulum conform to the
law that its period equals 2 piun * < TK of L / G similarly we might ask why in any free fall the velocity is propor proportional to the falling time as asserted by Galileo's law or why all planetary motions exhibit the Striking regularities expressed by Kepler's laws to many of these questions science offers an answer which interestingly has again the character of a deductive explanation a uniformity expressed by an empirical law is then explained by showing that it holds in virtue of or more precisely as a consequence of certain other more general laws or
of more fundamental and comprehensive theoretical principles for example the law for the simple pendulum as well as Galileo's and Kepler's laws can all be shown to be special consequences of the basic laws of mechanics and of gravitation similarly the optical laws of reflection and refraction and of Shadow formation can all be derived from the basic principles of the electromagnetic wave theory of light to develop theories which will thus explain the regularities expressed by empirical laws is one of the foremost objectives of scientific Endeavor and indeed many a scientist will hold that only where we can
offer explanatory theories have we attained genuine scientific understanding however that may be a sound theoretical explanation will certainly broaden as well as deepen our understanding of a given field of inquiry Newton's theory of motion and of gravitation for example broadened the scope of scientific understanding by accounting for a vastly wider range of phoma then do the previously established laws which the theory explains and the theory deepened our understanding not only in the sense of reducing all those other laws to one common system of underlying basic principles but also by showing that the previously accepted empirical
laws such as Kepler and galileos do not hold strictly but only in approximation for example Newton's principles imply that since every planet is subject to gravitational forces not only only from the Sun but also from the other planets the planets will not move in strict accordance with Kepler's laws but will show certain perturbations and Newton's Theory shows equally that the acceleration of Free Fall near the Earth is not strictly a constant as asserted by Galileo's law but changes with a distance of the falling body from the Earth's center of gravity all the explanatory laws and
theoretical principles I have mentioned so far have one important logical characteristic in common they are of strictly Universal form that is they assert the existence of certain unexceptional uniform connections for example between the volume the pressure and the temperature of a gas or between the temperature and the length of a copper Rod laws of strictly Universal form differ fundamentally from those of another type which during the past 100 years or so have steadily gained in scientific importance namely laws of probabilistic form in a nutshell the difference between the two is this a law of universal
form asserts that in all cases without exception when conditions of a specified kind C are realized a phenomenal on of a certain kind e occurs whereas a probabilistic law states that under condition C there is a statistical probability say R for the occurrence of e so that in the long run the proportion of cases of C that result in E will be R the laws stating the Half Lives of radioactive substances are of this kind for example to say that the half life of polonium is 3 minutes is to say that the probability for a
polonium atom to undergo radioactive decay within any given 3 minute interval is 1/2 this law can be used to explain why of a given initial amount of polonium only half is left after 3 minutes only 1/4 after 6 minutes and so on the basic principles of quantum the Theory provide another example of probabilistic laws and so do certain laws of genetics such as those which serve to explain the proportions of plants with white red and pink flowers that are found among the hybrids obtained by crossing a pure white flowed and a pure red flowed strain
explanations based on such probabilistic laws I will call probabilistic explanations because of the statistical character of the laws it involves a probabilistic explanation cannot show that the phenomenon to be explained was bound to occur it shows only that in view of the specified probabilistic laws and particular circumstances the phenomenon was to be expected with more or less high probability whereas a deductive explanation shows that given the truth of the explanatory information the occurrence of the phenomenon in question follows with deductive certainty but deductive and probabilistic explanation agree in their essential Reliance on covering laws both
explain a given phenomenon by showing that it occurs in conformance with such laws I think that this is indeed the common characteristic of all scientific explanation and more specifically that the scientific explanation of any empirical phenomenon is basically a covering law explanation of the deductive or of the probabilistic variety I will try to amplify and support this idea as I proceed it is sometimes said that an explanation should make us understand an unfamiliar or novel fact by reducing or assimilating it to facts with which we are already familiar but this idea surely does not adequately
characterize scientific ific explanation quite apart from the vagueness and subjectivity of the notion of familiarity here invoked this conception suggests first of all that familiar facts require no explanation but while in our everyday Affairs we might agree with this view science clearly does not indeed science has gone to Great Lengths in an effort to explain such familiar phenomena as the changing tides thunderstorm and rainbows the blue color of the sky the similarities between parents and their offspring our slips of pen tongue and memory and many others the point is strikingly illustrated by olber Paradox in
1826 the German astronomer Hinrich Alber pointed out that as a consequence of a few simple and extremely plausible assumptions including some standard laws of Optics and the hypothesis that the stars are distributed uniformly throughout the Universe the sky should appear very bright in all directions day and night thus the fact that it is dark at night which surely is as familiar as any observation was seen as posing a serious problem that called for an explanation a solution has recently been suggested on the basis of the COS ological theory of the expansion of the universe for
it can be shown that the uniform recession of distant sources of light would account for the fact that it is dark at night here then a very familiar fact is explained in terms of a theory that incorporates some unfamiliar and indeed quite extraordinary ideas this example also illustrates a second Point namely that rather than reduce the unfamiliar to The Familiar science will often do just the opposite it will explain familiar facts by powerful theoretical ideas which may strike us as quite unfamiliar and unintuitive but which account for a wide variety of phenomena and which are
well supported by the results of careful tests The View that to explain is to reduce the unfamiliar to the familiar can be deceptive in yet another way consider for example an explanation of a thunderstorm in terms of the wrath of an angered deity or consider a vitalistic account of some self-regulatory biological process as being governed by an inell or a vital force these explanatory attempts May well convey a sense of at homeness or familiarity with the Striking phenomena in question but they afford no genuine understanding of them vitalistic accounts for example do not tell us
in a general fashion under what conditions a vital force will manifest itself what specific form its manifestation will take and to what extent it will compensate for disturbing influences that have been exerted upon an organism thus vitalistic accounts do not indicate what occurrences are to be expected in this or that kind of situation consequently they cannot form the basis of any scientific explanation by contrast the explanation of planetary motion in terms of the Newtonian theory of gravitational attraction does tell us what gravitational forces will be exerted upon a given Planet by the Sun and by
other planets given their Mar masses and their distances and it tells us further what kind of motion is to be expected as a result of the action of those forces thus while both accounts invoke certain factors that cannot be directly observed one of them vital forces the other gravitational forces the latter has explanatory power while the former does not this is due precisely to the fact that gravitational forces are held to conform to specifically stated laws whereas vitalistic doctrines offer no laws governing vital forces thus it is laws that are essential for an explanation but
not the familiarity of the images and associations that the words gravitational force and vital force May conjure up in our minds the laws invoked in explaining a phenomenon also have predictive import they will predict first of all similar occurrences under similar circumstances for example capian motion for the planets of other stars and the laws may also predict new phenomena quite different from those they were invoked to explain for example the Newtonian laws at used to explain the capian uniformities of planetary motion also predict given the flattened shape of the Earth that the acceleration of Free
Fall is greater at the poles than at the equator indeed such predictions provide a means of testing the soundness of a given explanation and an explanation whose covering laws or theoretical principles fail on this score has to be abandoned such was the Fate for example of the explanation of combustion offered by the fistan theory no such fate can be Beall explanations in terms of myths or metaphors or vital forces since they do not tell us what to expect under any empirical conditions no empirical finding can ever discredit them but such absolute immunity from this confirmation
is not an asset but a fatal defect when judged by the goals of science for science seeks to establish an objectively testable and empirically well supported body of factual knowledge and account in terms of myths or metaphors however appealing it may be intuitively has no implications concerning empirical facts and thus has no logical bearing on them at all it is a Pudo explanation an explanation in appearance only I suggested earlier that the explanation of any empirical phenomenon involves its deductive or probabilistic subsumption under covering laws or theories now it is widely held that while this
might conceivably be the case in the Natural Sciences it surely is not so in other fields of research particularly in the psychological sociological and historical study of human thought and action according to this view an adequate explanation of an action requires not a specification of causes or covering laws but of motivating reasons these will include first the objectives the agent intended to achieve and secondly his beliefs about relevant empirical matters such as What alternative causes of action were open to him and what their probable consequences would be on one recent version of this view which
has its roots in collingwood's work an explanation of a given action must show that the action makes sense in the light of the agent's reasons or more precisely that it was the appropriate or reasonable thing to do Under the given circumstances which circumstances include especially the agent objectives and beliefs but leaving theide problem of standards of appropriateness here the explanatory information that a given action X was the appropriate thing to do Under the given circumstances does not as an explanation should entitle us to conclude hence it was to be expected that the agent would do
X to justify that conclusion we clearly need a further explanatory assumption roughly to the effect that the agent was the kind of person who will generally perform whatever action is appropriate in the situation in which he finds himself but an explanatory Assumption of this kind clearly makes a general claim and thus plays the role of a covering law though the law is of a somewhat peculiar kind it concerns the way a particular agent will generally act in a variety of circumstances the Oxford philosopher Gilbert riyle who has dealt in detail with the explanatory use of
sentences of this kind calls them law-like sentences thus it is true that psychologists and historians as well as people in their everyday dealings with each other will often invoke motivating reasons in order to explain human actions but this does not show that the explanations thus arrived at do not Pres suppose covering laws or at least covering law-like statements let me restate the reason briefly in more general terms whenever in an effort to explain a person's actions we attribute to him certain beliefs intentions moral standards character traits or the like then we assert by implication certain
law-like generalizations about the way the agent will regularly or or probably behave under various circumstances and it is on these generalizations that our explanation rests hence even the explanation of human behavior by reference to psychological characteristics and to motivating reasons is basically explanation by subsumption under covering laws this conclusion and the arguments that led us to it clearly do not imply a narrowly mechanistic view of man and his actions what our considerations suggest is rather this explanations in all areas of scientific inquiry share certain basic characteristics and deductive and probabilistic subsumption under covering laws constitute
modes of explanation that reach far beyond causal and mechanical explanation in particular subsumption undercovering generalizations is implicit also in those explanatory accounts that seek to exhibit the role played by conscious and unconscious motives and by ideas and ideals in shaping human decisions and actions and in affecting the course of man's history you have heard Professor Carl G hemple of Princeton University speak on scientific explanation next week for the 12th lecture in the philosophy of science series professor gral of the University of Pittsburgh will speak on space and time this is Frank Oliver inviting you to
join us again next week at this time for Forum the Arts and Sciences in mid-century America
