Methods of Induction and how they Work
For those who subscribe to the scientific method known as induction, science always begins with the direct observation of single facts on the ground that nothing else is observable, certainly not regularities. In its broadest sense, induction is any process that proceeds from statements made about some things of a certain kind to a conclusion or generalization about some or all the remaining things of that king. In brief, any rationale of induction would need to justify such statements as ‘All observed X’s are Y’s, so all X’s are Y’s.’ John Stuart Mill held that induction is the simple process of recording natural correlations (Morgan 13).
He further enumerated five different methods of induction, namely: (1) direct method of agreement, which affirms that if an occurrence in which the incident under examination have only a single condition in common, the situation in which alone all the cases in point concur is the reason (or result) of the said occurrence; (2) method of difference which affirms that if an occurrence in which the event under examination happens and an occasion in which it does not happen have every situation in common except for one, that one happening only in the past; the situation in which alone the occurrences diverge is the result, or the reason, or a crucial part of the source of the incident; (3) joint method of agreement and difference, in where the methods of agreement and difference is used together in one extended line of argument, is a procedure for eliminating items from a list of suspected necessary and sufficient causes; (4) method of residues, which is to seize away from a happening what is recognized by preceding inductions to be the result of certain precursory conditions, and the remainder is the consequence of the antecedents left behind; and (5) method of concomitant variations that is used specifically in cases in which the phenomenon under investigation and the conditions causally related to it are simultaneous.
Methods of Deduction and how they Work
Deduction is a conclusion or inference derived by reasoning from formal premises or propositions. A main difference between induction and deduction is that in deduction the truth of all the premises is held to guarantee the outcome or conclusion, whereas in induction this is not so. Deduction is not only an approach to research method and methodology; it is also an approach to explanation. In the philosophy of science this mode of explanation is more properly called ‘nomological-deductive explanation’. A number of assumptions underpin the deductive process. The first of these is that discrete and measurable variables can be identified. The second assumption concerns the place and role of values in research. Though values are accepted as a necessary part of the act of identifying the variable and its consequent gradations, they do not subsequently act as distorting or biasing moments in the deductive process.
Essentially, the methods of deduction might be divided into the following: first, concepts, in where the decision to which concepts represents important aspects of the theory or problem under investigation are to be made. However, since concepts are abstract they are not readily observable, and therefore the asserted relationships between concepts provided by the theory are not open to empirical testing until these abstractions are translated into observables or indicators – that is, they have to be operationalized; second, through the operationalization of a concept it becomes defined in such a way that rules are laid down for making observations and determining when an instance of the concept has empirically occurred; third, by creating rules for making observations we are making a clear definition of what it is we are going to observe. In this we create indicators, or measures, which represent empirically observable instances and occurrences of the concepts under investigation; fourth, the process of operationalization enables the construction of clear and specific instructions about what and how to observe.
Empiricism and its Relationship to Induction and Deduction
The most important source of modern-day empiricism and positivism is not Francis Bacon but rather the eighteenth-century British philosopher David Hume. Empiricism is a doctrine insisting that valid knowledge can only be based on sensory experience – in short, on the observed features of the social world as they present themselves to our senses. It has been associated with a ‘correspondence’ conception of truth in that theory and concepts are thought to ‘correspond’ in some direct (almost ‘photographic’) manner to the ‘facts’ or ‘empirical data’. It means that all knowledge must be sensed to be real, with the use of experimental methods to validate a theory; faith alone — knowing that it is true because you believe it to be so – is an insufficient basis for explaining a phenomenon or as a foundation for knowledge.
Empiricism, as noted above, is the view that the only grounds for justified belief are those that rest ultimately on observation. Based upon the work of philosophers such as Hume and John Locke, its central premise is that science must be based on a phenomenalist nominalism, that is to say the notion that only statements about phenomena which can be directly experienced can count as knowledge and that any statements that do not refer to independent atomized objects cannot be granted the status of justified knowledge. Unlike rationalism, according to Emile Durkheim, empiricism does not assume that there are any ‘internal connections’ or ‘logical relations’ among things that ‘allow one to think of some with the help of and as a function of the others’ (qtd. in Sobrinho 68).
Empiricists believe that science can rest on bedrock of such pure observation, and from this bedrock can be established, by induction, the entire scientific structure. Put simply, basic beliefs, warranted by direct perception, provide basis for induction so that we can move away from a very narrow foundation for knowledge to much wider inductive generalizations.
Among the most important of these to philosophers are induction and deduction. What this means is that philosophers sometimes use deductive argumentation in the service of an empiricist epistemology, a classic example being Hume’s attitude toward his epistemological principle that every simple idea is a copy of a previous impression. Similarly, philosophers sometimes use inductive arguments in support of rationalist epistomologies, Cohen’s example being Whewell’s historical arguments to justify his Kantian epistemology.
For empiricism in particular this meant that one had to start from indubitable factual propositions from which, by gradual valid induction, one could arrive at theories of ever higher order. The growth of knowledge was an accumulation of eternal truths: of facts and ‘inductive generalizations’. This theory of ‘inductive ascent’ was the methodological message of Bacon, Newton and – in a modified form – even of Whewell. Critical practice demolished the classical idea of valid content-increasing inferences in both mathematics and science, and separated valid ‘deduction’ from invalid ‘informal proof and ‘induction’. Only inferences which did not increase logical content came to be regarded as valid. This was the end of the logic of justification of classical empiricism. Its logic of discovery was first shaken by Kant and Whewell, then crushed by Duhem, and finally replaced by a new theory of the growth of knowledge by Popper.
Rationalism and its Relationship to Induction and Deduction
Rationalism is the view that knowledge is possible without experience. It puts its faith in priori knowledge – knowledge independent of experience and which is the result of various forms of argument and reasoning. The assumption is that reason provides us with clear and distinct ideas and guides us to conclusions that we can draw from these ideas. It has often been associated with a ‘coherence’ conception of truth in which the internal relationships between concepts in a theory are thought to be decisive in shaping the terms in which we understand the phenomena to which the theory refers. Rationalism was challenged by empiricists to explicate the origins of simple ideas. Such an explanation, the argument was made, was the whole point of science. On the surface it might seem that this mirrors exactly the debate about induction versus deduction with empiricism being concordant with induction and rationalism having more in common with deductive methods.
However, it is not as simple as this for it does seem that most approaches to ‘empirical theorizing’ have stressed empiricism as the most secure and fundamental basis of knowledge. A fundamental problem of the approach is how to identify clear and distinct ideas. The implication of this system of thought is that the roots, or beginning of truth is an absolute and pure truth, which is used to build or synthesize further truths or proofs. The foundational truths re derived by intuition and further truths arrived at by deductive reasoning, using the basic ideas as foundation. Influenced by rationalism, some contemporary empiricists prefer to draw a line of distinction between primary and secondary induction and believe that whereas the former, although really inductive, is not as such justifiable, when supplemented by the latter (which is deductive) becomes so.
Method Embraced by Modern Science
Although rational deduction and empirical induction remain with us today, both forms of inquiry suffer from numerous shortcomings. In particular, these three are that the assumption of rationalism that the premises in a deductive argument are true without resorting to sensory observation, the assumption of empiricism that all knowledge is based on sensory experience and the assumption of empiricism that the limited past and present observations provide a basis to predict future observations alone (qtd. in Miller and Brewer 68). This last shortcoming was one of the major contributions of Hum in specifying the problem of induction.
Modern science has, as a result, consisted of interaction between empiricism and rationalism, between induction and deduction; any effort to draw the line on either side clashes with authentic scientific performance. This is agreed upon by Whitehead’s remarks on the nature of scientific enquiry when he explicitly points out that science is never capable of proceeding very far by the methods of pure deduction or of pure induction. Instead, the method used in contemporary practice is always a synthesis of the two modes. One might surmise that a priori sciences (deductive) operate exclusively with deduction, and that the a posteriori sciences (inductive) work only with induction, to the exclusion of all deduction.
Combined induction, deduction, empiricism and rationalism express the full force of scientific method for resolving the greatest contemporary complication. Induction alone, or any of the other three alone, are equally incompetent to the great problems of the scientific community. To speak of ‘deduction versus induction’ is to misconstrue the scope of these two areas of logic. If induction is in need of justification, so is deduction. Ultimately, the issue of the relation between deduction and induction, empiricism and rationalism must be understood as discourses which are potentially open to each other’s influence and which are capable of producing positive and constructive interactions – rather than incompatible and incommensurable epistemological starting points.
The knowledge of numerous instances of something happening, i.e. an inductive process, may cause the development of a theory from which are deduced possibly refutable hypotheses, which are then tried out empirically. An important aspect of this is the theories of truth upon which rationalism and empiricism are based. While recognizing that both have something to offer, it is important to avoid either the extreme coherence or the extreme correspondence theories since they both lead to exaggerated and erroneous claims about the better hold on the truth (greater explanatory power). In order to look for the most adequate and powerful forms of explanation, adaptive theory draws upon both theories and occupies the intermediate ground between them in an effort to transcend the limitations of both.
Miller, Robert and John Brewer. The A-Z of Social Research. Bonhill Street, London: Sagel Publications, Ltd., 2003.
Morgan, Michael, ed. Classics of Moral and Political Theory. Indianapolis, Indiana: Hackett Publishing Company, Inc., 2005.
Sobrinho, Blasco José. Signs, Solidarities and Sociology. Lanham, Maryland: Rowman ; Littlefield Publishers, Inc., 2001.