In the last blog post, the term “objectivity” was regularly used, however, the definition of this term was not as clearly defined as it could have been. In this post I will attempt to explain how one can still be justified in saying that something is “objective”, while at the same time acknowledging the problems with the term “objective”.
When the term “objective” is used, often times it is referring to something that is believed to be ontologically true. However, in science and other areas of life it is essential to analyse what we mean when we say that something is true scientifically. Evidence for whether a statement is considered to be objective or not, depends on evidence for that statement.
In the philosophy of science, the ravens paradox (Hempel’s Ravens) is worth analysing. Consider the proposition that “all ravens are black”. This can be understood to be an objective fact because to date no one has observed anything that is a none black raven. When this is the case, does observing anything that is not black and not a raven become evidence for the proposition that “all ravens are black”? At the moment I am writing this blog post on a grey laptop, does the fact that my laptop is not black and not a raven, count as evidence for the proposition that “all ravens are black”?
This might sound banal, but often times scientists' will base their views on whether something is true on a series of studies. Some can extend this logic to philosophical questions, for instance one of the main arguments against God’s existence is that there is no evidence for God, some scientists’ postulate that with every scientific discovery the hypothesis for God becomes weaker. The raven paradox reveals a contradiction between inductive logic and intuition.
The paradox reveals problems with evaluating evidence in cognitive science. In psychology, if P happens to be greater than 0.5 on a variety of different occasions, one might think that after a series of studies reveal that P>.05 this confirms that the hypothesis is ontologically true. This approach is not unproblematic, because it could very well be the case that certain research methodologies contradict one another. Alternatively there may be flaws with the individual studies or even the set P value. Also journals are less likely to publish articles based on experiments with a null hypothesis, so publicly it may appear as if a series of tests confirm a certain hypothesis when the reverse could be true.
Hopefully by analysing solutions to the raven paradox we can arrive at a better definition of what we mean we say that something is objectively true or scientifically true.
Hempel’s solution is to make a similar comparison to the proposition: “all sodium salts burn yellow”. If one holds a piece of ice to fire instead of salt and it fails to burn yellow, does this confirm hypothesis? According to Hempel, this can only count as evidence in favour for the first proposition if one avoids any reference to previous knowledge. If one thinks about the solution to the raven paradox as a scientific hypothesis, can we be justified in saying that a proposition is ontologically true if we don’t have prior knowledge?
Maher sought to accept the conclusion of the paradox but refine it. A non-raven confirms that “all ravens are black” because the fact that the object is not a raven removes the possibility that the object is a counter example to the generalisation. It also reduces the probability that unobserved objects are ravens.
The Bayesian-Carnap solution states that it’s not the case that observing anything that is non-black and non-raven enforces the hypothesis, but instead that the prevalence of black ravens means that “all ravens are blacks”. This proposition is also slightly modified because it would consider the existence of something that is non-black and non-raven (such as a grey laptop) to be part of the same proposition as “all ravens are black”. Thus, it helps to reaffirm the proposition that all ravens in existence are black.
The role of background knowledge was mentioned in the Carnap solution, another problem with why a scientific truth can not necessarily be said to be ontologically true is because
scientific theories inevitably rely on knowledge from previous theories. It is also impossible to test a scientific theory in it’s entirely (see Duhem-Quine Thesis). Also one cannot have all of the background knowledge regarding a theory, so theories are susceptible to change and modification based on new information.
If anything is to be referred to as being “objectively true”, one should not infer that this is the same as saying that something is ontologically true. Saying that something is “objective” should be understood as saying something is believed to be true based on the existing data. Perhaps this approach is more instrumentalist or anti-realist, but when we can’t be sure if something is ontologically true it only makes sense to embrace this definition of “objectivity”.