Return to Summary Page ..................... Return to Previous Section ..................... Go to Next Section

2. Knowledge of the world is acquired from experience under the interpretation of reason.

2.1 By reason, I mean the deduction of conclusions from initial premises (propositions) using the rules of logic. To provide knowledge of the world, these premises must be based on experience.

2.1.1 A distinction is to be drawn between a priori propositions & self evident propositions. A priori propositions are prior to all observation of the (or any possible) world, describing how the world (allegedly) must be. Self evident propositions are those most fundamental propositions for which no further justification can be found. A self evident proposition is a response to the world, whereas an a priori proposition is a prescription for the world though often masquerading as a response to it. An individual's set of self-evident propositions forms (to quote Richard Rorty) his final vocabulary. A priori propositions that are alleged to refer to matters of fact, rather than being mere tautologies, are to be treated with suspicion. They may only express the prejudices of the proposer. Alleged proofs of what must be the case in the world, based only on a priori propositions but not on actual observation, are all false. Therefore, attempted proofs of the necessary existence or necessary attributes of God (such as the Ontological Argument) which are based solely on a priori propositions are invalid. Other arguments for the existence of God (such as the Cosmological Argument) do not fall foul of this thesis, being responses to observation (eg. to the appearance of design in the universe). Such arguments may, of course, be invalid for other reasons, and further consideration will be given to them later in this paper. It is possible to explain the meaning of the word God by adding various attributes (such as omniscience, omnipotence, omnipresence etc.), but this is all part of the definition of a term. It is not a proof that any entity in the world answering to the name God exists. It may seem possible to repudiate all self-evident propositions and so dissolve into vicious regress. However, it need not be so in practice where there is a genuine desire for communication. Those wishing to communicate must agree some minimal set of common premises. While such a minimal set of premises does not exist of necessity (ie. it cannot be proved beyond all contradiction to exist) in practice it does exist, because this minimal set of propositions is a response to the world which, by definition, exists.

2.1.2 There is only one logic or method of reasoning. Since logic is basic to argument and to thought, it is not possible to for or argue against a system of logic other than that commonly received. It is also difficult to imagine what an alternative system of logic would be like. Work has been done on developing so-called deviant or alternative logics (eg. those that deny the law of excluded middle), but has not so far borne fruit. In Quine's phrase, those who redefine basic logical terms simply "change the subject". Since there is only one valid system of logic, the expressions human logic or human reason, in supposed opposition to divine logic or divine reason, are confused notions. However, the propositions accepted as self-evident by individuals or groups may be false & based on prejudice. Correct reasoning based on incorrect premises will usually lead to false statements about the world. A distinction needs to be drawn between logic as such and its formalism. The formalism of logic is capable of development (eg. from the propositional calculus to modal logic) and is still being so developed. Also, alternative formalisms are possible. However, logic, in the sense of rules of thought is as old as its applications. Logic is not a response to the world. I assume that logic as normally understood is a prescription for any possible world. Certainly, the distinction often drawn between synthetic and analytic propositions presupposes this.

2.1.3 Systems of knowledge, such as mathematics, that are based on axioms and subsequent reasonings, do not constitute knowledge of the world, but may be used as models or pictures of the world. For example, as is well known, Euclidean geometry is an inadequate partial model of the world in the presence of matter (assuming the General Theory of Relativity to be true : in any case, the existence of this theory demonstrates that space is not necessarily Euclidean). Even ordinary arithmetic is not necessarily true of the world, as is shown by the existence of many abstract algebras. It is possible to imagine worlds in which the propositions of arithmetic did not hold (eg. addition could always be modulo 24, as the hours of a clock). in support of the contention that the rules of logic are necessarily true of the world, while those of arithmetic are not, it is sufficient to note that it has been proved, contra Whitehead and Russell, that Number theory cannot be reduced to logic. Godel's Incompleteness Theorem demonstrates (in more mathematical/logical language) that any axiomatisation of number theory contains theorems that can be stated but not proved within the formalism. Since mathematics (and, in particular, number theory) cannot be reduced to logic, it follows that any & all elements of a mathematical model may be false of the world while logic is true of it.

2.1.4 Similarly, as has been explained above, a priori theology is not necessarily true of the world, and its "proofs" are unsound when they assert allegedly necessary statements of fact. Theologies are models of the world, and whether or not they have any application outside themselves can only be determined by inspection of the world.

2.1.5 We must distinguish between sense and reference. An idea may have sense, in that we understand what it means, but have no reference, in that there is nothing corresponding to it in the world. Classic examples of ideas with sense but no reference are unicorns and fairies (though there is more sense to the concept of a unicorn than to that of a fairy, in that we can better analyse what a unicorn would be like if it existed (by analogy with a horse) than we can a fairy). It is important to avoid ideas with reference but no sense, ie. confused notions based on unanalysed, or incorrectly analysed, experience. For example, objects may be named ostensively, by pointing to them and giving them a name. However, if that name is more than a label, and has a meaning of its own that is intended to describe the object, the name will be senseless if the description is wrong. For instance, if I imagine I see a ghost, then the term ghost has sense, but no reference, on the reasonable assumption that ghosts do not exist. On the other hand, what I point to, and refer to as a ghost, has reference in that it exists (be it a patch of moonlight on the wall, or whatever) but has no sense because the data has been analysed incorrectly.

2.1.6 The idea of a game is also useful in picturing or modelling the world. We define the rules of the game, and play according to these rules, hence revealing the inner structure of the game. This inner structure may have properties that were not apparent before we started to play the game (eg. chess). Theology may be treated as a game with its own rules. Theologians must beware lest their game has sense but no reference. The statements of theology must constantly be referred to the world.

2.2 By experience, I mean observation of the world, whether studied or casual.

2.2.1 Since I am part of the world, experience of the world includes the observation of my own thoughts & feelings. I refer to this as subjective experience, in opposition to objective experience which applies to objects outside myself (paradoxically including my own body). The above remark should not be taken to mean that I believe in a disembodied mind or soul. I am simply distinguishing between sense experience and other experiences (eg. dreams, thoughts, feelings, pleasure, pain etc.). Nor should the term subjective necessarily be understood in the pejorative sense of less reliable.

2.3 Objective experience may be considered to be built up of atomic sense perceptions of the form "I see a red patch now". However, in order for the individual to perceive anything significant, these atomic perceptions need to be combined and analysed. That is, sense perception has to be interpreted before it can be understood. Reason has its part to play in converting sense data into experience even before it is applied to experiences to form a body of knowledge about the world.

2.3.1 The process of the infant's perceptual development involves the assembly of complexes of sense perceptions into recognisable objects and experienced situations. These developments apply to each of the five senses (sight, touch, taste, smell & hearing). From want of conclusive evidence to the contrary, I take it that there are no senses apart from the normal five. Also, there is no such faculty as extra-sensory perception. Other so-called senses (such as the religious sense, aesthetic sense, moral sense etc.) are compounds of these five senses and emotional reactions to ideas.

2.3.2 In childhood & adult life, experiences & situations are "recognised" as complexes. There is a tendency to translate raw sense data into an experience that is expected. Hence, it is easy for our senses to be deceived, for we are not perceiving the raw sense data, but only the interpreted image.

2.3.3 What is experienced depends on what we understand the world to contain. To an extent, we perceive what we expect to perceive. Conversely, we may fail to perceive what we do not expect to perceive. While perception is generally conservative, ie. unusual situations are converted into familiar ones, an expectation of an unusual event occurring may stimulate an hallucinatory perception. For example, we are more likely to see a ghost on the landing if we have just seen a ghost film.

2.3.4 The more time we have to consider an experience, the less likely we are to misinterpret it, because we have a greater opportunity to re-evaluate the situation. Our minds (or "brains", assuming the truth of the mind-brain identity theory) continue to search for interpretations of experiences until a "fit" is found. In the case of leisurely perceptions, the search may continue until the best available fit is obtained. In the case of urgent situations, the first approximate fit may be accepted as the correct interpretation, though we may subsequently persuade ourselves that we could not have perceived what we thought we had if the alleged perception contradicts our understanding of the world. Hence, if we believe in flying saucers (or nodding statues, or angels) we may generate them as possible "fits" to a situation, whereas if we don't, they will be low on our list of possibilities : we will tend to screen them out.

2.4 A distinction is to be made between experience & experiments. In an experiment, the conditions are controlled so that the experimenter can determine whether the (pre-selected) proposition p or the proposition not-p is true of the world.

2.4.1 This definition conforms to the Galilean concept of experiment, as being the more fruitful, rather than the Baconian, in which no hypothesis is being tested, but the world is simply being observed to see how it performs under controlled conditions. An experiment should be so designed that any appropriately equipped & serious person can repeat it with the same result.

2.4.2 Statements about the world that cannot be tested by experiment are to be treated with suspicion, unless they follow logically from statements or complexes of statements that are subject to experimentation.

2.4.3 Some form of verification principle is essential, namely, that unverifiable statements have almost certainly no reference. This is because in a contingent world there are infinitely many possibilities. I sympathise with the basic philosophical methods of the logical positivists. This does not imply, however, that I agree with all their underlying aims or with all of the consequences of their approach. By logical positivism I mean the view that statements and putative entities whose accuracy or existence cannot be verified by possible sense experience are meaningless. However, the foundation principle of logical positivism, that the meaning of any statement is its method of verification, is probably false. This would imply (as in the previous proposition) that unverifiable statements are meaningless rather than having sense but, possibly, no reference. I accept Popper's criticism of logical positivism's verification principle : that no amount of positive evidence can verify a general law, though one contrary observation may falsify it. I accept his "Conjectures & Refutations" approach whereby scientific hypotheses are not arrived at inductively but are well-informed guesses (conjectures) which may be refuted by experience but cannot be proved by experience. I take it, however, that statements of a particular nature (ie. once off statements about the world, eg. "Napoleon lost the battle of Waterloo") are equally verifiable or falsifiable. While I would not go so far as to deny meaning to unverifiable or unfalsifiable statements, I would insist that such statements be treated with caution. Both the logical positivists' verification principle & Popper's falsification doctrine are intended as principles of demarcation. The verification principle is intended as a criterion of meaning, to distinguish sense from nonsense. Popper's falsification doctrine is intended to distinguish science from pseudo-science. I am interested in both principles as criteria for evaluating truth claims. In what follows, I will use the term undecidable for unverifiable or unfalsifiable to avoid tedium.

2.4.4 A distinction needs to be drawn between statements that are undecidable in principle and those merely so in practise. An example of a statement that is undecidable in principle is "there exists a universe parallel to ours with which we cannot interact". Examples of statements that are undecidable in practise are of several kinds. For example, open statements such as "gravity is always attractive" and areas reflecting current technical incompetence such as "the centre of the Earth is made of iron" etc. We note that an open statement such as that above is unverifiable in principle but only undecidable in practise, because finding two mutually repulsive masses would falsify the general law.

2.4.5 Statements that are undecidable in principle are more suspect than those that are merely undecidable in practise.

2.4.6 The allegedly unverifiable statement "there exists a transcendent God" is only obviously so if such a God is supposed never to become immanent. The Judeo-Christian God, who is understood not to be contained by the universe, may be considered to be transcendent. however, since he is also understood to get involved with the world, he is knowable and therefore his existence & putative attributes are verifiable to the extent to which he becomes involved. Speculations about what God is like "in himself" (ie. in his transcendence), whether Trinitarian or otherwise, are unverifiable in principle & are therefore best avoided. We might ask how we might come to know God in his transcendence if all we know of him is gleaned from his immanent interactions with us. It could only be because we were told of it. Even then, it is difficult to see how we could accept the divine statement except on trust.

2.5 The chief & most successful means of acquiring systematic knowledge of the world is by means of science, which may be spoken of both as an activity and as a body of knowledge, the collected results of the application of the scientific method.

2.5.1 As an activity, science is a method of asking and answering questions about the world. Any discipline that asks questions that are capable of being answered by controlled observation (ie. by experiment) may be termed a science. Hence, the historical & social sciences may be so categorised along with the physical sciences. A naive view of the application of scientific method to some subset of experience runs somewhat as follows :-

a). Gather & classify as many relevant types of data in this subset of experience as are available.

b). Develop a set of internally consistent models to account for this data.

c). Select the best fit model which satisfies the extra demands of conformity with other established laws. This may lead either to a new law or (more frequently) to an explanation of the phenomena within the framework of established laws.

d). If no adequate model exists, examine the established laws to see whether they require modification. This is a very rare event, leading to what Thomas Kuhn refers to as a paradigm shift.

e). Once an adequate model has been developed, test it against new data not used in the development of the model. This further testing will establish firstly the truth and secondly the scope and boundaries of the model. The most fruitful application of the scientific method has been in the physical sciences.

2.5.2 Science, considered as a body of knowledge, is a collection of interrelated theories about the world. A theory is a model or picture of some limited domain of experience that is deemed to have been satisfactorily confirmed by experiment. Prior to verification by experiment (or rather to exposure to possible falsification by experiment) a model with a measure of explanatory power, otherwise known as an hypothesis. In fact, as Popper has shown, no general law can ever strictly be verified. However, our confidence in it increases in proportion to the number of tests it survives without falsification. A theory that is quite adequate in one domain may be inadequate in another. The classic example is Newton's theory of dynamics, which is only a good approximation to the world at speeds small compared to that of light.

2.5.3 There are domains of experience of the physical world that are not susceptible to direct experimentation. Such are all descriptions of the past or of the future, and in particular of the history & future of the universe as a whole, which is the subject of cosmology. However, because of the finite speed of light, all of our observation of the universe is of past events and all observation of deep space is effectively the archaeology of the remote past of the universe. Another example of a domain of science not susceptible to direct experimentation is the study of the development of living beings, of which the Darwinian Theory of Evolution is an attempted explanation. Strictly speaking, the theories in such areas have the status of hypotheses only, since they cannot be directly confirmed or refuted by experiment. However, such hypotheses are genuinely scientific because they have consequences that may be confirmed or refuted by experiment. We must be on our guard against hypotheses that can accommodate themselves to any new experience. An hypothesis that will not allow itself to be refuted by any experience is simply metaphysical. Examples of such theories (suggested by Popper) are Marxism and Freudian psychoanalysis.

2.5.4 A general hypothesis that is ubiquitous in science is that of uniformitarianism. That is, that the laws of physics are taken to be constant with respect to both space & time. The reasons for this hypothesis are twofold. Firstly, it seems to be borne out by the limited corner of spacetime available for our analysis. Secondly, it is the only workable assumption we can adopt. We would get nowhere in science if we were to assume that the laws of physics vary capriciously whenever they are out of the range of investigation. Where exceptions are entertained, for instance in Guth's inflationary model of the universe, this is because of boundary conditions at which we have no right to expect the normal laws of physics to apply. Hawking's attempts to describe a universe without boundary conditions is a neater solution which, if coherent, is to be preferred.

2.5.5 The scientific method is paradigmatic for any activity that claims to provide knowledge of the world. Its primacy is demonstrated by its practical success.

2.6 Historical knowledge is a form of experience at second (or more remote) hand. The claims of historical science as to what happened in the past are validated by application of the reason to historical records.

2.6.1 The availability of historical experiences for contemporary analysis is dependent on reliable records of them being preserved. Where the preserved account is written, it is important that the primitive inscripturation should have been a faithful record of what took place. In general, this will mean that the records were written by independent witnesses close to the time when the events are alleged to have taken place. Also, where the primitive record was made on a medium now lost or decayed, it is important that numerous early independent copies of the document were made and that some have been preserved. In certain societies, oral tradition may supplement or replace a written one. However, such traditions are likely to change over time with the society itself, as the society reworks its history in the light of its present state or to justify its aspirations. Hence, oral traditions are usually only reliable in a static society. If a transient medium (whether oral or written) was used to record historical experiences, the transmission of these experiences in their original form depends on the reverence with which these traditions were held and on the conservatism of the society preserving the traditions. If the society was both creative and conservative, a custom of pseudonymous writings may have developed, in which new ideas were projected back into the normative past. Such was the case during the inter-testamental period in Israel and in the sub-apostolic period in Christianity, though whether this pseudonymity applied to any of the canonical Scriptures (eg. to Daniel) is a question much debated. Apart from evaluating the written & oral testimonies to past experience, it is also essential to take into account the copious evidence of the workings of former societies supplied by archaeological research (eg. artifacts, buildings etc.). Finally, we can obtain evidence of the non-social past from the sciences of geology & palaeontology and by the extrapolation back in time of presently active processes. The precis of the historical method given in this section is restricted to the methods applicable to Biblical research & to investigations into other ancient societies. The methods & means applicable to modern history, such as the use & evaluation of pictorial & aural records by way of film & soundtrack, are ignored.

2.6.2 The statements of history can be evaluated by reasonings based on common experience. These reasonings apply both to the evaluation of specific evidence adduced in favour of historical claims and to the general probability of such claims.

2.6.3 The following items, at least, must be taken into account when evaluating specific written evidence adduced to support historical claims :-

a). The remoteness of the period.

b). The number & independence of the sources.

c). The contemporary publicity of the episode recorded.

d). The belief-set of the society in which the event in question occurred or by which it was recorded.

e). The general political character of the society, ie. whether it was a totalitarian society or one in which free expression of opinion was allowed.

f). The purpose for which the record was made.

g). The character of the witnesses, as in judicial proceedings, in so far as their characters may themselves be determined by reliable testimony or deduction.

h). The educational, literary & cultural background of the witnesses.

i). The general bias of the witnesses (eg. as displayed in peripheral matters).

2.6.4 When evaluating archaeological evidence, we must take into account the following criteria :-

a). The climatic & other conditions that have influenced the spectrum of objects available for contemporary study.

b). The number of similar artifacts preserved. Care must be taken to avoid extrapolating general theories from statistically insignificant evidence.

c). The general character of the culture under study. For instance, did it have an interest in, and the power to achieve, the suppression of evidence ?

d). Our general understanding of the particular culture under consideration. We are likely to misinterpret the use of artifacts, buildings etc. if we are largely ignorant of the aims & ethos of the culture. Archaeology is a cyclical, self-reinforcing (or self-correcting) study in which detailed findings are used to build up successive approximations which are themselves refined by new findings.

2.6.5 Geological and palaeontological evidence must be subjected to the same statistical tests as archaeological evidence. Where evidence is statistically insignificant, it should be treated with extreme caution. However, there may be occasions when the lack of evidence may not be statistically significant. For instance, when a rapid transition is presupposed, the existence of many transitional forms is not to be expected. This is notoriously the case in the theory of evolution. While the lack of evidence is to be expected, this lack does put pressure on the theory, which may be considered unfalsifiable and hence metaphysical.

2.6.6 With respect to the general probability of an alleged historical event, the following factors, at least, are relevant :-

a). The more significant or unusual the event, whether political, social or physical, the more testimony is required to establish it.

b). In the case of political or social events, the more significant the event, the more likely it was to have been generally discussed, whether verbally or in writing, and the more likely permanent records were to have been made.

c). A general lack of recorded evidence for a significant event is evidence against its having occurred.

d). It is only possible to combine our knowledge of the world into a coordinated world-view if we do not allow our world-view to be disturbed by every new piece of alleged data. Hence, it is a good general policy to reject poorly substantiated events as fictitious, especially if they do not fit in to the model of the world that we have constructed from more reliable data.

2.7 Innate knowledge is not available, though innate abilities & propensities most probably are (cf. Chomsky & language acquisition). Hence, the mind at birth is a tabula rasa with respect to knowledge but not with respect to the ability to acquire those skills which are subsequently essential for the acquisition of knowledge.

2.7.1 The specific promptings of conscience are socially acquired, rather than being innate, as is demonstrated by the various workings of conscience in different societies. However, the ability to acquire & operate within a moral code would seem to be innate & to be essential for social animals that are not totally dominated by instinctive behaviour. The appearance of a measure of commonality of moral codes amongst different societies may be explained from the observation that certain practises (or restraints therefrom) are essential for the preservation of any society. Societies that are deviant in key areas have a tendency to die out or disintegrate. Familiarity & social conditioning give the appearance of inevitability to practises that are essentially arbitrary.

2.8 There will always be areas of experience that are unexplained within any current understanding of the world. The two extreme viewpoints below are best avoided.

2.8.1 Firstly, we wish to avoid using God to explain the gaps in our knowledge, bringing him on as a Deus ex Machina.

2.8.2 Secondly, we wish to avoid forcing recalcitrant knowledge into currently understood patterns. There is nothing wrong with a genuine admission of ignorance, however regrettable that ignorance may be.

2.9 In the above, we have been concentrating on the first-hand acquisition of knowledge through personal experience and personal reasoning. In practise, because the range of possible experience is so wide, the majority of any individual's knowledge of the world is acquired from books, teachers or other secondary media.

2.9.1 We will resume this problematical issue of secondary knowledge when we come to consider world views in the next section.

© Theo Todman 1992 - 2000.
Go to Commentary on this Section
Return to Previous Section
Go to Next Section
Return to Summary Page