*/f
2
What Kind of Science
is Social Science?
Classical Responses
[There] is a controversy which for mote than half a century Uns split not 'only logicians and methodologies but also social »dentists into two schooh of thought. One of these holds that the methods of the natural sciences which ha*e brought ■bělil such magnificent results are the only ones and that they alone, therefore, have lo be applied in their entirely to the study of human affairs ... The other school of thought feci» that there is a basic difference in the structuic of the social world and the world of nature. This feeling has led to the other extreme, namely the conclusion that the methods of the social sciences are iota lorh different from those of the natural sciences ... It has been maintained that (he social sciences arc ideographic, characterized by individualizing conceptualization and seeking singular assertory propositions, whereas the natural sciences aie nomothetic, characterized by generalizing conceptualization and seeking general apodictic propositions. The latter have to deal with constant relations of magnitude which can be measured and can perform experiments, whereas neither measurement nor experiment u practicable in the social sciences. In general, it is held that the natural sdenecs have to deal with matcnal objects and processes, the sodal sciences, however, with psychological and intellectual ones and thai, therefore, the method of the former consist in explaining, and that of the latter in understanding
A. Schutz. 'Concept and Theory Fonnation in the Sodal Sdenecs'
Introduction
The controversy which Schulz identified in this quotation has raged for at Icasl | hundred years. While some progress has been made in achieving a better understanding of the nature of the different views, and some proposals for alternative positions to bridge the differences have been offered, the issue of the extent to which social life can be studied in the same way as nature continues to be the central one in the philosophy of social science (Keat and Urcy 1975; Bhaskar^979; Held and Thompson 1989).
In order to answer the question "Can the methods of the natural seien» es be applied, fr» flip 8QcÍpl.BCi«nre»s?'. another question must also be answered: 'What are the methods of the natural sciences?' For the past fifty years

12    Science and Social Science
thaw has been as much if not more controversy over this question than there has been over Ihe firsl one. Many philosophies of sciiínee have been proposed ami debated centring on issues such as ihr nature and importance Ol observation, when obsorviidon should occur in the process ot developing scientific knowledge, the appropriate form of logic to be used in constructing theories, the role of theories themselves in this process, the structure of theories, and the extent to which scientists work with open minds or arc constrained by the beliefs, values and orthodox practices of the community of scientists to which they belong. To answer the first question it will be necessary to deal with the second as well.
It is important to note that the reference lo "methods' in these questions docs not refer to the actuul techniques of observation, data gathering or data analysis, but rather to the logic or strategy of enquiry, lo the processes by which knowledge is generated and justified. The techniques of data gathering and analysis used in the various disciplines are related to the particular nature of their subject matters. While some techniques may be used in more (hail one discipline, ci1 i he use til' parti» ulai statistical tests. the study of chemical structures is a very different activity from the study o| social stun lures
It is not possible to answer the first question with a simple 'Yes' or 'No'. First, more than one method has been claimed to be appropriate in the natural sciences and, secondly, both 'Yes' and 'No' answers have been qualified in various ways. Eleven different answers to the question will be explored, six classical positions in this chapter and five contemporary positions in the next chapter. The first of the classical positions, and the one against which all other positions are reactions, is known as Positivism
the onl> straight Yes' answer; u advocate that ..II sciences, whelhei natural 01 social, should use the same method. The second answer is that of Negativism - a straight-out 'No'; it claims (hat a science of the social is impossible. The third answer is 'Yes and No\ referred to here as Histoncism; it advocates a modified form of the method of Ihe natural sciences proposed by Positivism as being appropriate for the social sciences, one in which prediction based on past trends is the primary concern. The fourth answer, another 'Yes and No', is known as Critical Rationalism', it also argues for the use of the same methods, but rejects the view of science associated With Positivism in favour of a different method lor natural science. The fifth answer, from Classical Hermeneutics, is 'No'; it claims that ihe natural science aim of explanation is not relevant to the social sciences and it is concerned with interpretation, parlicularly of lexis. The sixth, anoiher 'No' answer, is Interprelivism; it rejects the methods of the natural sciences as appropriate for the social sciences arguing that, because of the qualitative dilleienecs in then subject matteis, a diflcrent kind »it scieniilic method is required.
This brief overview of the classical positions makes it clear that it is not possible to give a simple answer to the question. A vast and complex array of philosophical positions and arguments will be encountered in the review
»>l
Classical Responses   13
lHldl1                     u           »s  oViv   het»MliKla.1»Mlsl.H.l(»,.s1d»lat,onofthe
Classical Responses
Positivism
The first answer is a straightforward 'Yes'. It is based on a position-known as Naturalism, the belief that there can be a natural scientific study pf people nul society, the doctrine known as the 'unity of scientific m ľ I hod', U is argued that in spite of the differences in subject matter ol the various scientific disciplines, both natural and social, the same method or logic ol explanation can be used, although each science musl elaborate these in a way appropriate to its objects of enquiry (Popper 1961; von Wright 1971; Kolakowski 1072). The philosopher John Stewart Mill (1879) adopted this position as a way of rescuing the social (or moral) sciences from what he regarded as an unsatisfactory Mate. He bc)icycd_tha_t all scientific explanations have fundamentally the same logical structure. A more recent philosopher, Kail Poppet, adopted a mmlai position.
I do not intend to assert that there are no differences whatever between the methods of the theoretical sciences of nature and of society; such differences clemly exist, even between the various natural science» themselves, .is w< II as between the various social sciences . .. Hul I agree with Comic and Mill - and many others ... - that the methods in the two ftekh «e fundament.illv the same. (Popper l%l: 130M)
Popper, as shall be seen, had his particular view of the scientific method which differed from that advocated by Mill and the other early supporters of Naturalism. The fact that there are differences in subject matter is not seen to be a problém: According to this standpoint, the phenomena of human subjectivity, of volition and will, do not offer any particular barn-ersTcHne treatment of social conduct as an "object" on a par with objects in the natural world' (Giddens 1974: 3-4).
In his definition of 'science', Brailhwniie has included
all the natural sciences, physical and biological, and also such parts o( psychology and the social sciences (anthropology, sociology, economics) as are concerned with empirical subject-matter* .. This sense of the word scicncc-corresponds pretty closely with the most frequent modern use ol the word ... ; it is synonymous with natural science' if man [sic) is included with nature (Braithwaite 1953: \)
28
14    Science and Social Science
This thesis of Naturalism is a Central claim of Positivism^.d philosophy of the natural sciences which, in its various forms, has been both the dominant doctrine and the object of extensive criticism. In the nineteenth century. Positivism was not merely a philosophy of science; it expressed a more general worldview which lauded the achievements of science. While the term was invented by Comtc (1830). one Of the founding fathers of sociology, he did not formulate the doctrine.
Numerous attempts have since been made lo identify the central tenets of Positivism (sec, for example, Abbagano 1967; von Wright 1971; Kolakowski 1972; Giedymin 1975; Hacking 1983; Stockman 1983). Jn addition lyjtu' dicsis_of. the 'unity oLmcthoiľ, the.following rules, or beheIs ;ir(.**geiici:illv accepted as characterizing PoSitivisCÜ
Phenomenalism This rule asserts the uniqueness of experience as the onlv reliable EasisTor scíénííiic knowledge. 'Wc ire entitled to record only that which is actually manifested in experience* (Kolakowski 1972: 11). That which is to count as knowledge must be based on experience, on what ui observer can perceive by his or her senses. 'The positive tradition gets its character from an insistence on the epistemic primacy of direct observation. The senses give us our sole direct acquaintance with the world, our only source of unvarnished news' (Hollis 1977: 44). This perception must be achieved without the subjective activity of the cognitive process; it must be 'pure experience' with an empty consciousness. Hence, it is a passive model of knowledge. 'An episode of scientific discovery begins with the plain and unembroidercd evidence of the senses - with innocent, unprejudiced observation' (Medewar 1969b: 147).
Nominalism The rule of nominulism asserts that any abstract concepts that are used in scientific explanations must also be derived from experience; metaphysical notions about which it is not possible to make any observations, have no legitimate existence except as names or words. As Hindess (1977:16) has stated it. '(pjositivism asserts the claims of experience as the ultimate foundation of human knowledge and denies the possibility of meaningful discourse concerning supersensible objects.' For example, the concept of 'God' cannot be rcgarded_as scientific as it is not possible to observe God, and statements such as 'God exists' or 'God does not exist' are meaningless because no observational evidence can have any bearing on them. A development of this rule is the belief that the language used to describq.pbscrvations must be uncontaminated by any theoretical notions and that the statements in this language can be readily established as true or false by reference to 'reality'. It is a theoretically neutral observation language in which the descriptive terms correspond to real objects. Hence, descriptive terms which do not belong to this privileged language. i.e. theoretical terms which arc unobservablc. must either be able to be translated into observables or they will be regarded as meaningless.
*/»
Classical Responses    15
Atomism: The. objects of experience,.of observation, are regarded as discrete," independent, atomic impressions of events which constitute the
ultimate and fundamental elements of the world. In so far as these atomic impressions are formed "into generalizations, they do not refer to abstract objects in the world, only regularities among atomic events i Harré lv)""ill
Central Laws   Scientific theories are regarded as a set of highly general. law lik ľ statements, and establishing such general laws is the aim of science These scientific laws summarize observations by specifying simple relations or constant conjunctions between phenomena. Explanations arc achieved by subsuming individual cases under appropriate laws. These liiws ure general in scope in that they cover a broad range of observations and tfičořc universal in form m that they apply, without exception, across time and space: 'Positivism pivots on the Humean theory of causal law». viz. that laws are or depend upon constant conjunctions of atomistic events or states of affairs, interpreted as the objects of actual or possible experience. This theory itself follows ineluctably from the requirements that knowledge be certain and given-in-experience' (Bhaskar 1979: 158).
Value Judgements and Normative Statements _ This rule requires a scpa-ration of 'facts' and 'values' and denies values as having the status ol knowledge: 'we arc obliged to reject the assumption of values as characteristics of the world for they are not discoverable in the same way as the only kind of knowledge worthy of the name' (Kolakowski 1972: 17). Giddens hus expressed this rule as the idea that judgments of value have no empirical content of a sort which renders them accessible to any tests of their "validity" in the light of experience' (1974: 3).
I lacking (1983: 41-57) has identified a similar set of six key ideas which are associated with a positivistic approach to the natural sciences.
1    There is an emphasis on verification: the truth or falsity of statements about the world can be settled in some way, they can predict observable states of affairs and are capable of conflicting with observable
evidence.
2    Pro*obstrvation: what we can see. feel, touch etc, provides the best
foundation for non-mathematical knowledge.
3    Ami-cause; there is no causality in nature, only constant conjunctions between events, such that events of one kind are followed by events of
another kind.
4    Downplaying explanations: if all we have are regularities between types of events, then explanation is nothing more than locating an event within a wider-ranging regularity; explanations do not provide deep
answers to why questions.
5    Anti-thtorettcal entities: given me preceding ideas, unobscrvable entities may not be offered as causes because reality is what can be observed, and observed regularities are all that can be verified.
16    Science and Social Science
u Against inrUi/'/ivM, \ lín- positivisf argues ih.il nuď-i.ilvJi" propositions u n observable entitles, causes and deep explanations are derived from metaphysical notions and must be avoided.
According lo Halfpenny (I982).it is possible to identify twelve varieties oi Positivism. However, for the present purposes, these can be reduced to three (following Outhwanc 1987a). The first brand of Positivism was formulated by Comre as nn alternative to theological and metaphysical ways Of understanding the world. All scientific knowledge is based on causa] laws derived from observation, and all sciences are unified in a hierarchy oi relaiccUÍcvels, building on mathematics at the lowest level, followed h\ isiioimmy', physics, chemistry, biology and, finally, sociology. However. I omle believed in the existence  qt a social  realm   independent of the i    ilnie> of the sciences at lower levels in the hierarchy, and governed by laws that cannot therefore be reduced to the laws of the othei sciences; he i ejected rcduc.tionis.nl.
The second brand of Positivism, known as I ogical Positivism, was founded in Vienna m (he 1920s. The catch cry of these philosophers was thai any concept or proposition that does not correspond to some state-of-alfairs. i.e. which cannot be verified hy experience, is regarded as mean-iiiľjcss (the 'phenomenalism' rule). At "the same lime, it is argued that the concepts and propositions of the higher level sciences can be reduced to Ihoae of the lower ones. In other words, they adopted the reductionist position that the propositions of the social sciences could ultimately be analysed down to those of physics.
Hie third variety, which was derived from the second and is sometimes referred to as tlie 'standard view' in the philosophy of science, dominated (he BřJgllsh-Spealíing World iftar the Second World War Its fundamental tenet is that all sciences, including (he social sciences, are concerned with developing explanations in the form of universal laws or generalizations. Any phenomenon is explained by demonstrating thai it is a specific case ol some such law. These laws are of the form of 'constant conjunctions' between events, or in the case of the social sciences, statistical correlations Oi icgularities (the 'general law' rule).
At its most general, positivism ii a theory of the nature, oniiii-compctence and unity of science, hi its most radical shape ir stipulates that the only valid kind uf í nun-analytic)1 knowledge is scientific, (hat such knowledge consists in (he description ■>( i he invaiianl patterns, ihe co exigence in space and succession over (ime. «Í observable phenomena ... lis naturalistu místence on the unity of science and scientist ic disavowal of any knowledge apart Irniii science induce its .iveismu m metaphysics, insistence upon a strict value/
1 Analytic stiitcmcnts are Iruc l»y detlnlllan leg. "Black cats arc hla.k), or ,'iin lie tautological («,| I llhei Mack eats arc Walk i»r they «re not black*), ti is sellconirndiclory u* deny either of ihes* kinds of statement*. Synthetic knowledge, on the othci hand, nukes a claim about IhC world (e.g. 'Black cats tue fierce'), and can be denied without self coniradKiion.
Ml
Classical Responses    17
faci dichotomy and tendency lo hisloncist conlidence in the bwviubry oi scientifically mediated progress. (Bhaskar 1986: 226)
The various brands of Positivism, while differing in some detail, have a particular view ol the methods ol Ihe natural sciences. Other responses lo the question of whether the methods of the natural sciences can be used in the social sciences have not accepted this view. It is therefore useful to distinguish between Naturalism, as a positive response lo the question, and Ihe specific features lhal have come to be identified with Positivism as a philosophy of science, Ihus making it possible to adopl a naturalism based on a non-positivisi view of science (Keat and Urry 1975: 2).
It was through the woik ol t unite (19711) ami Dutkhcim (1964) that Positivism was introduced into sociology, forms of Positivism have dominated sociology, particularly in the decades immediately following the Second World War, and continue to do so today in disciplines such as psychology and economies In the past fifteen yeais, Positivism has been the subject of much discussion within sociology (sec, for example. Giddens 1974; Fay 1975; Keat and Urry 1975; Adorno et al. 1976; Benton 1977; Hindcss *1977; Halfpenny 1982; Bryant 1985). Some examples of how Durkheim advocated the application of the methods of the natural sei ences in sociology will be discussed in chapter 5.
Negativism
Ihe proponents ol this answer may accept a Positivisl view of the natural sciences bul argue lli.il Ihe methods of the nahnal sciences cannot be applied"to the social sciences. The argumenls loi  lins assertion come in various forms and Rave been usefully summarized bv Popper (1961: S 34).:
1     While the so-called 'laws of nature'* are assumed to apply through out space and time, the regularities in social life are time and space specific. Social uniformities cannot be generalized because they change from one historical period to another and across cultures. The reason for this, it is claimed, is that social uniformities arc Ihe result of human activity - they are socially constructed - and can therefore be changed by human activity. 'For social uniformities are not laws of nature, but man-made [sic|; and although they may be said to depend on human nature, they do so because human nature has the power to aller and, perhaps, to control ihern" (Popper 1961: 7-K>
2    The use of the experimental method is based on the assumption thai similar things will happen in similar circumstances. By artificially isolating
' ľ'or similar reviews nee Na«el (I'Kil) and Runeimun (1%'i).
' these argumenls, like much ol llic popular di/ussion ol ihr niiUiic of science, assume ihat Ihe 'laws of nature* are ical, lhal ihey have an existence independent ol the scientist. Popper, however, regarded such laws a« Ihe scientist's inventions which .ire designed to provide an understanding of the natural work!.
96
10