Into the Great Uncertain
Logical Positivism’s Acceptance of Uncertain Knowledge

C.A. Postlethwaite

Overview

An important aim of science is to produce knowledge of the external world. In order to advocate this aim of science, Logical Positivism (LP) had to face External World Skepticism (EWS), the view that we cannot know anything about the external world with any certainty. Under this paper’s reading, LP inherited EWS from its early precursor, Classical Empiricism (CE), as the main obstacle for any non-moderate empiricism to be fully supportive of science. This paper first offers a brief account of CE and then of LP especially in regards to these matters. Afterwards, it evaluates whether LP is an improvement over CE, in the sense of being more supportive of science while still maintaining an empiricist program. I posit that LP not only continued CE’s empiricist program but also proposed that EWS could be overcome by pursuing a logical way to account for less than certain claims and knowledge. Finally, I conclude that LP’s move to embrace uncertainty in science together with its intuition to develop an inductive system of logic, can be seen as an improvement over CE’s positions with respect to science; even if the problems that later arose for LP had no solutions to be found, and search for them has for the most part ceased.   

Classical Empiricism

Very generally, empiricism encompasses distinct views that tend to favor experience as the basis for human knowledge. I follow Peter Godfrey-Smith[1] in delimiting CE to the philosophical empiricist movement of the seventeenth and eighteenth centuries and represented by John Locke (1632-1703), George Berkeley (1685-1753) and David Hume (1711-1776). CE, under this reading, will lead to EWS. It is important to note that CE cannot be seen to have considered the posterior strategy of Immanuel Kant (1724-1804) of “reigning-in empiricism”[2] nor can it be seen to have knowledge of phenomenalism’s assertion that “the external world is just a concept of a patterned collection of sensations”[3].  In this sense, CE is exempt of Kant’s proposal to tackle EWS and of phenomenalism’s barefaced acceptance of it.

In accordance to empiricism’s main thrust, CE’s views sustain that experience is the only source of knowledge. Experience itself, in turn, is solely acquired through the senses. For CE, the mind starts off as a completely “blank tablet”[4]. The mind is thus totally empty except for received sensations and whatever mental activity can be applied to them. (P1) Sensations, on this view, are the only content that our minds have access to. CE asserts that said content, sensations, can be manipulated by the mind in order to form beliefs. Our minds can conceptualize the received sensations, establish relationships between them and detect additional instantiations of sensations in further experience.

The problem that arises for this reported view of CE is of major concern for science. (P2) If the only content that our minds can access are our sensations, then our beliefs and knowledge can only be about our sensations. (P3) If our beliefs and knowledge can only be about our sensations, then our beliefs and knowledge cannot be guaranteed to be about anything external to our sensations. (P4) The external world, if it exists, is external to our sensations. (C5) Thus under CE, our beliefs and knowledge cannot be guaranteed to be about the external world. This would be a hard blow for science indeed, since one of its main pursuits is to provide knowledge about the external world. But this pursuit seems closed-off if CE is successful as described above and we are interested only in truths about the world that can be logically guaranteed. There is another potential skeptical problem that can be harmful for science and attributed to some forms of CE. It is called inductive skepticism but it will not be part of this paper’s concerns.  

Logical Positivism[5]

Again, I will follow PGS in identifying LP only as the initial part of a movement that can be seen to have attenuated its claims and ambitions soon after 1940. LP surfaced in the 1920’s, around 150 years after CE. In the meantime, attempts had been made towards a more sober brand of empiricism, though the very sobriety put these attempts “in danger of losing their distinctive nature as empiricism”.[6] LP’s proponents wanted to return to a strong empiricist view by positing that everything we know about the world can be known only through experience. But unlike CE, LP was not at all skeptical about the possibility of knowledge of the external world.

For LP science’s claims were surely about the external world, and common sense could not deny science’s practical efficacy and advances in everyday life. What was needed, according to LP, was a system that explained and validated science’s undeniable success. By doing so, LP sought to help science along its pursuit of knowledge. LP was also influenced by advances in formal logic and the philosophy of language. With science and logic as its guiding principles, LP thought it found a way to settle all of philosophy’s problems. From LP’s perspective, every problem of philosophy was either solvable under its program or else meaningless.

LP posited that claims that were presumably about the world could be verified by observation, i.e. experience. LP followed Kant in calling these claims synthetic. Any factual claim (synthetic claim) that could not be verified was considered meaningless by LP. The distinction between verifiable and unverifiable claims demarcated the phenomena for which we could acquire knowledge. Only synthetic claims that were verifiable could be said to mean anything at all. This ensured that all knowledge about the purported external world would be empirical, since only experience could justify claims that were about the external world. This is one of the two central ideas of LP, it is called the Verificationist Theory of Meaning (VTM). A short way of stating VTM is that a sentence about the world has meaning if and only if we have a possible way to observe or test its truth.

There was another sort of claim that LP conceded to. These claims do not look to experience for their truth or falsity, they are true or false depending only on the meanings of the terms in sentences themselves. The truth of the claim “all bachelors are unmarried”, for example, depends only on the relation between the terms in the statement. These types of statements are not verified by observation and they are not about the world. They are called analytical propositions. LP admitted analytical propositions could be true but only vacuously so, since they just show “our conventional decision to use symbols in a particular way”.[7] These sorts of claims are not important for what will be advanced in this paper. It suffices to say that the Analytical/Synthetic distinction is the second central claim of LP.

For the interests of this paper, the crucial move that LP deployed concerns LP’s position on the testability of claims about the external world. Since science attempts to understand the external world, science depends wholly on observation and not just on the relations between meanings of words. Our scientific knowledge can be seen as true claims about the world. Analytic claims are not about the world, so they are not the focus of science. Claims about the world are justified by observations. But observations always carry a degree of error. Observations can never produce claims that are certainly true. Since science relies on uncertain observations for the truth of its claims, then the claims of science are all uncertain.

LP pointed out that if the nature of scientific enquiry was to deal with and produce uncertain claims, then deductive arguments were not compatible with science. The philosophy of science had to look for a way for less than certain premises to support or confirm less than certain conclusions. Arguments with these characteristics are not deductive arguments, rather they are called inductive.     

Notes

[1] Godfrey-Smith, Peter (2003:19) Theory and Reality, The University of Chicago Press.

[2] Strategy offered by Kant after 1870. See: Lacey, Alan (2005: 242) ‘Empiricism,’ in Ted Honderich Oxford Companion to Philosophy, Oxford University Press. Also: Allison, Henry E. (2005: 466) ‘Kant, Immanuel,’ in Ted Honderich Oxford Companion to Philosophy, Oxford University Press.

[3] Godfrey-Smith, Peter (2003:20) Theory and Reality, The University of Chicago Press.

[4] Lacey, Alan (2005: 242) ‘Empiricism,’ in Ted Honderich Oxford Companion to Philosophy, Oxford University Press.

[5] Because of length requirements, I am not providing an adequate depiction of LP. I have chosen to present only the issues pertinent to the evaluation proposed in the beginning of this paper.

[6] Lacey, Alan (2005: 242) ‘Empiricism,’ in Ted Honderich Oxford Companion to Philosophy, Oxford University Press.

[7] Godfrey-Smith, Peter (2003:26) Theory and Reality, The University of Chicago Press.