Coherence or Correspondence: The Schlick – Hempel Debate

In the Vienna Circle there occurred a debate about how to characterize truth. On one side was Moritz Schlick, an adherent of the correspondence theory. The representatives on the other side were Otto Neurath and Carl Hempel. Schlick accused Neurath of holding a coherence theory of truth. Neurath could not distinguish science from fantasy, according to Schlick. In this paper, Schlick’s objection to Neurath will be examined, as will Hempel’s attempt to dodge Schlick’s criticism.

Schlick believed that science needed a foundation to ground the practice of observation and confirmation, and that serves as a test for figuring out when sentences match the facts (Schlick 209-227, 215-16, 220-222). The sentences that are relevant to these tests are constructed from demonstrative indexicals such as “this”, “here”, and “now”.  An example would be an hypothesis predicting a certain color experience if a certain surface is observed, and a person having that experience. She would pick out what she experiences by saying “blue here now” while pointing at the surface (Schlick 225). Those utterances, called “affirmations”, form a set of unrevisable sentences that constitute the test for when we are in contact with the facts. These affirmations would be so thin that they could not be mistaken in the way that a more detailed description of the situation could be, and so would be unrevisable. A description that picked out the time, date, and other factors without only employing demonstrative indexicals has room for a mismatch between itself and what it describes.

According to Schlick, a coherence theory of truth characterizes truth as the mutual agreement of statements within a system (Schlick 214). According to both parties engaged in this debate, coherence is merely the lack of contradiction within a set of sentences. Schlick accuses Neurath of being a coherentist because Neurath believed that the notion of comparing sentences with “facts” or “reality” was unintelligible at best, and metaphysics at worst (Hempel 50-51).

The reason why Hempel and Neurath’s views were Schlick’s targets is because they represented a move away from the correspondence theory. They disagreed over what’s required for confirmation; rather than a set of unrevisable sentences against which hypotheses are tested, there is a shifting set of sentences that is relevant for particular acts of confirmation, but also stands in need of confirmation by other statements in other circumstances. That shifting set of sentences constitutes the background knowledge held by scientists, and as such it becomes the resource from which they draw to craft hypotheses and entailed predictions in need of confirmation or disconfirmation. Science is like a ship that needs to be altered at sea, where some parts serve as foundations but can also be called into question at a later date (Hempel 54).

Since all sentences are open to revision, Schlick’s criterion for truth cannot exist. There aren’t unrevisable sentences that serve as the point from which other sentence alterations occur, and this is why Schlick has a big problem with coherentism. If there is no criterion from which we can distinguish legitimate inquiry from practices like myth making, then there is no distinction between science and fantasy. Science is supposed to be aimed at getting us in contact with facts, so it cannot be concerned with truth as Schlick sees it if Hempel’s and Neurath’s views are correct (Schlick 212-216). All that’s required for a true set of sentences is for those sentences to be consistent. So, truth becomes relativized to systems of sentences lacking contradiction.

Hempel and Neurath have sufficient resources to defend their theory from Schlick. One way is to show that science can be distinguished from myth making empirically. It is possible to observe which cultural group assents to sentences when they are predicted by hypotheses formed from common background knowledge. When these predictions are confirmed, this particular group assents to the sentence(s) entailed by that confirmation, and when the predictions are disconfirmed, the group dissents from those sentences. “Scientific truth” becomes consistency within the language of science (Hempel 57). Scientists indeed aim at truth, but “truth” means that they aim to make their background set of sentences consistent in light of confirmations made through observations. Scientists and myth makers are methodologically distinguishable. In other words, the institution of science and its practice of forming hypotheses against the background knowledge scientists share, which involves confirming or disconfirming the predictions made by those hypotheses is its distinguishing feature.

Schlick’s charge that there is no distinction between science and fantasy turns out to be false. Science can be characterized by the methods and practices of a particular cultural group, and those methods are distinct from myth making by virtue of their goal – forming hypotheses from background knowledge, and then making predictions entailed by those hypotheses, and engaging in observation to confirm or disconfirm those predictions. Scientists are engaged making observations and revising their beliefs in light of recalcitrant observations, whereas the people we call myth makers are spinning yarns for entertainment. The myth makers are not worried about the role of observation and confirmation. What makes science so special is its methods, and those can be empirically discovered by observing scientists engaged in their daily activities.

The response from Hempel would not satisfy Schlick. He would ask what makes these methods of this particular group so important that we should privilege them over some other group’s. What exactly grounds the observations we make if there is no privileged class of sentences? Couldn’t the set of sentences that constitutes the background eventually be revised so much so that it resembles the sentences that constitute a fairytale?

Hempel and Neurath should dig in their heels. The description of the institution of science is enough to provide grounds for privileging the sets of sentences to which scientists assent. The methods involved in forming fairytales would exclude them from being the background of this particular group. Furthermore, they have three reasons not to embrace Schlick’s criterion. One reason is that “facts” as objects of correspondence are sentence-like slices of reality which sentences aim to represent. But representation requires metaphysical notions like “intentionality”.

The second reason is that affirmations are not the sorts of things that are intersubjectively accessible. Occasions of affirmation cannot be repeated because of their indexical nature; each repetition is actually a different affirmation, so they cannot be considered publicly accessible. There is no way for somebody to repeat an affirmation such that it can serve as a public criterion for confirmation. Since these affirmations are private, they serve no purpose for scientists. Schlick’s criterion has no behavioral difference in the practices of scientists, so there is no reason to think that they employ it in their daily activities.

The last reason is that Schlick’s criterion can either serve no inferential role in scientific theories, or loses its status as unrevisable. The instances of seeing blue aren’t the things that play an inferential role within a set of sentences, since only sentences can stand in entailment relations with other sentences. So affirmations have to be the things which stand in inferential relations with sentences. But in this case, it becomes a comparison between two or more sentences, and not experiences and sentences. We end up using sentence comparisons as the litmus test for truth, and this is exactly what Neurath affirms.

So, Hempel and Neurath preserved the distinction between science and fantasy. While Schlick’s was an historically important criticism in light of the persecution faced by many members of the Vienna Circle, it ends up toothless in light of the resources at Hempel’s and Neurath’s disposal.


Works Cited

Hempel, Carl G. “The Logical Positivist’s Theory of Truth.” Analysis 2.4 (1935): 49-59. Print.

Schlick, Mortiz. “The Foundation of Knowledge.” Logical Positivism. Ed. A. J. Ayer. New York: Free, 1959. Print.