According to the account of Auguste Comte’s Positivist philosophy written by the illustrious John Stuart Mill, Comte agreed with those bold philosophers who asserted that we cannot assume that causation actually exists in the Universe. All we know of reality is its surface features. What lies beneath those, we have no way of fathoming. All we know of the Universe is relative, not absolute– relative in that everything we know or think we know about reality comes first through the filters of our five senses. And we are far from being able to state the reason for why a thing happens ( called by some “final cause”).
As for what some have termed an “efficient cause”– or an event which solely and directly gives birth to a specific, following event that we call an “effect”– we cannot even know this, for these two events may occur one after the other without fail in all times and places of the Universe, and yet, the first may not necessarily by the “cause” of the second. We can only say that the two events are highly correlated– perhaps we can even say that they are invariably linked in an invariable order– that the first will never occur without the second, nor the second without the first– and yet, we must hold back from pronouncing that the first is therefore the cause of the second. Such a statement is beyond our knowledge and experience.
Mill tells us that the closest Comte would come to calling things causes was to refer to them as “invariable antecedents.” Nevertheless, because some chains of events are invariable, we are justifiable in ascribing certain laws to Nature. In other words, we don’t have to understand Nature to describe or predict it. For this reason, it seems clear to me that Comte’s Positivist approach has a fair claim to paternity when it comes to what is “modern” about modern physics. The modern physicist is much more concerned about describing and predicting events (even if he settles for probabilities) than explaining them. And just to be clear, Comte adds that these invariable relations are NOT subject to intervention by any deity or devil.
Comte demands that any hypothesis must adequately explain the fact and be susceptible to direct proof. Furthermore, a well-made hypothesis must be stated in such a way that, if it is true– that is, if it accords precisely with the observable facts– than the negation of the hypothesis can NOT accord with the facts. For example, if we hypothesize that, for a specifically defined situation, “what goes up must always come down,” then for that hypothesis to be acceptable, it must be shown also that it’s opposite, “what goes up must NOT always come down,” is false (for the specific situation defined). Otherwise, you don’t have a valid hypothesis according to Comte.
Additionally, a scientist should refrain from using analogies to explain things he does not understand. For instance, Comte probably would not have approved of the one-time description of the atom as a miniaturized Solar System with electrons orbiting around a nucleus. This assertion was never proven (and has in fact fallen out of favor), but it was attractive to scientists of the day because it was analogous to things they could see at the macro level.
However, even Comte did not claim to banish the supernatural altogether from the Universe. As Mill writes, “the laws of Nature cannot account for their own origins.” Whatever existed before Nature was, by definition, supernatural, something occurring outside Natural Laws.