To explain literally means to ‘straighten out’ in the sense of removing irregularities. It is a process of making something intelligible or saying why certain things are as they are; In short, an explanation provides answers to why questions: “Why did the child behave in that way?” It is not always easy to make a distinction between explanation and description. For example, consider the following statement: “Newborns when given appropriate postural support in an appropriate state, reach for, but not grasp, a suspended object”. This statement can be taken as a description of when newborns show reaching movements or it can be used to explain as to why newborns can be induced, under certain conditions, to produce such movements. An acceptable explanation is to be found in the functional context, namely, they produce reaching movements when supported in a semi-upright position, but not when lying down. There are a number of scientific explanations that relate specifically to behavior, and which can be summarized as follows (the first two of which can be distinguished based on the strength of the explanatory link between the explanans and the explanandum):
* deductive (or nomological) explanations (as in the deductive-nomological model): in reality, they are arguments in which a statement describing an event to be explained (the explanandum) is the conclusion (C) and a set of law-like statements (L1 to Ln), together with particular sentences (I1 to In) describing antecedent or correlated circumstances, are the premises (the explanans) purporting to do the explaining. More appropriately termed explanatory arguments, they assume that the truth of C follows logically or deductively from L and I together. For example, the infant cried (C) because she had not been fed for hours (I) and unfed infants always cry (L). Such arguments are sometimes distinguished in terms of production and retrodiction.
* probabilistic explanations: the truth of C cannot be logically inferred from I and L. The most one can infer is that it is more probable that C is true rather than false. Put another way, the explanandum follows as a probable consequence, rather than a hypothetical certainty, from the explanans. For example, this child has cerebral palsy (C) because he suffered brain damage at birth (I), and 40% of all such children have cerebral palsy (L). Together I and L only make a child’s cerebral palsy a probable outcome as it does not follow logically from them that he will have the disorder. Many explanations in the developmental sciences are of this type. It is important to recognize the limitations of probabilistic explanations. One example is that they do not assert that the occurrence of an event is inevitably accompanied by the occurrence of some other event. Rather, they hold that in a sufficiently long series of trials, the occurrence of one event is accompanied by that of a second one with an invariable relative frequency. Such explanations are not causal laws. Such laws may be deduced from statistical laws if suitable assumptions can be made about the statistical distribution of the initial conditions for the application of particular causal laws. This deductive step is based on the assumption that statistical laws (or more correctly stochastic processes) are the consequences of some underlying deterministic order. This assumption, brought about by the stochastic revolution created by quantum mechanics, treats chance (or chaos) as a deterministic phenomenon in which events arise from random interactions constrained by hidden parameters. Thus, order can be produced from disorder through the operation of certain constraints that are inherent in many physical processes.
Another way of identifying scientific explanations refers to the concepts recruited in the explanans. Such concepts can be causal, functional, purposive (or motivational), teleological or genetic in origin. Thus:
* causal explanations: in such explanations, C describes some event and a statement in I accounts for its cause. Some law-like statement provides the basis for inferring if the first event occurs or has occurred, then a second will or would have. For example, the infant cried (C) because he was hungry (I) and all hungry babies cry (L). In this essentialist view of causation, it is taken for granted that that factor X will cause an affect under all circumstances. In other words, X completes the set of sufficient conditions for the occurrence of Y. This view has virtually been abandoned in every branch of science since the advent of quantum mechanics except for very restricted circumstances that would hardly apply to the study of ontogenetic development in healthy children (e.g., Trisomy 21 always give rise to Down’s syndrome). A more appropriate causal explanation of development is systemic causality.
* functional explanations: answer questions such as “What is such-and-such thing for?” (e.g., what is the selective advantage of this behavior?). Typical of explanations in evolutionary biology, it is not an explanation of why something exists. Ideally, the study should combine causal and functional explanations, as the selective advantages of developing abilities can never be adequately understood apart from their causal mechanisms. Conversely, functional explanations lead to a greater understanding of the design features of a particular ability and thereby a clearer insight into the causal mechanisms involved.
* purposive explanations: perhaps the oldest of world views on behavior, it should not be confused with functional explanations as they explain some action or decision in terms of an agent’s intentions, motives, purposes, aims, preferences, fears etc. Often treated as non-scientific in that the explanans refers to mental events or states that are beyond the purview of scientific explanation, such explanations have been re-admitted to the study of child development, largely through the application of information-processing models. More especially though, purposive explanations have become established fixtures in research on the development of pre-verbal communication. However, they continue to be controversial and problematic, largely because they propose sufficient conditions (e.g., innate physical knowledge) whose origins remain obscure.
* teleological explanations: in the past, they have been used to as a general category to cover both functional and purposive explanations. Unlike the other two though, they endeavor to account for the directness of ontogenetic development in which species-characteristic end states can be achieved through a variety of pathways. Thus, teleological systems demonstrate equifinality, a redundancy or availability of functional equivalents that form the vary basis of adaptation in both ontogeny and phylogeny. They are so immanent in nature that the scientific status of teleological explanations can never be satisfactorily superseded by causal, mechanistic explanations of living organisms.
* genetic explanations: accounts of sequences of events leading up to the occurrence or existence of a fact to be explained, which in fact amount to descriptions rather than explanations. Much of the theorizing on motor development in past has consisted of genetic descriptions of sequences of explicanda in search of appropriate explanantia. At most, they were probabilistic explanations of antecedent-consequent relationships.
See Causality, Causal determinism, Constraint, Chaos, Deductive-nomological model, Description, Descriptive explanations, Equifinality, Explanans and explanandum, Implicate order, Law, Laws of nature, Law-like statement, Prediction and explanation, Prodiction and retrodiction, Quantum mechanics, Stochasticity, Systemic causality