understanding charles s. peirce's 'abduction and induction'

Peirce’s Abduction and Induction aims to make clear multiple forms of inference that may not be initially clear. He begins by stating that all knowledge rests on observable facts. That is, only when you observe something can you actually start to begin making sense of it – only then can we start bringing the idea to the third degree of clarity. Peirce defines a hypothesis as not just a supposition about something, but any other truths that may come from the fact that we have observed. In other words, a hypothesis is built off of the idea that observed facts (and the propositions that come along with these observations), can also be applicable to other scenarios. For example, Peirce mentions that a hypothesis is not only the idea that a man may be a Catholic priest because of observations we can make about his attire, but also includes the fact that perhaps we can identify priests from other religions based on the clothes that they wear.

In this paper, Peirce focuses on testable hypotheses. He goes on to explain that an abduction is the starting or entertaining of a hypothesis, and adopting that hypothesis as worthy of further testing through the scientific method. He believes that abduction is a form of inference because it can be classified using logical reasoning: C is observed, and if hypothesis A were true then C would follow, so there is reason to believe that A might be true. This final line of reasoning tells us that a possible hypothesis is worth pursuing, and this outlines the process of abduction.

Peirce goes on to outline his definition of induction, which consists of realizing that if a hypothesis is true, then if we design an experiment and subject it to certain conditions, our observed results will be consistent with this hypothesis. The level of confidence that we attribute to the hypothesis after these experimental results are obtained, is called induction. Peirce brings up an example where someone may be led to believe that more female children are born to Black parents than to white parents. In order to test this hypothesis, we can look at a large sample of census data, and say that there was relatively a large population of female children born to Black parents when compared with white parents. Because of the observed ratio we have found through experimentation, we can say that our hypothesis is valid, and is a strong induction. Clearly, induction involves a great deal of statistical reasoning, as it pertains to ratios and proportions of the observations we find during the experimentation process.

However, there are certainly situations in which the observed phenomena cannot be quantified in such a precise manner. Let us go back to the example of a Catholic priest. Peirce mentions that Catholic priests are all very familiar with the Italian pronunciation of Latin, so if we pronounce something with this accent (which most people unfamiliar with the pronunciation would not understand) and the man we hypothesize to be a Catholic priest is able to understand it, we may be able to infer that he is a Catholic priest. However, how do we know what weight this data has? That is, how do we know that the man is not just familiar with the Italian pronunciation for a reason unrelated to being a Catholic priest? This is an example of a weak confirmation, or an abductory induction. This kind of induction relies less on quantitative data and the understanding that characteristics cannot be counted statistically or mathematically – we can only guess how much weight we should lend to this kind of data.

Peirce now moves on to the idea of deduction, which is largely the idea of making conclusions from accepted premises rather than data that we are going to observe through experimentation. Peirce brings up an example in which we hypothesize that a man believes that the Pope is never wrong. It follows logically that if the man does not believe in immaculate conception (which the Pope obviously believes to be true), then the man does not think that the Pope is never wrong, and our hypothesis has been falsified.

These are four types of inference which are extremely important to the scientific method. First, we come up with our hypothesis through abduction. Then, we reason about the premises through deduction. Finally, we carry out our experiment and come to a conclusion about the strength of the hypothesis through induction or abductory induction.