What’s a fact, anyway?

by Neil Rickert

It is common for discussions of knowledge and science to emphasize the role of evidence, where “evidence” usually refers to facts.  Typically, facts are said to be the starting point.  But what, exactly, is a fact?  That’s what I want to examine here.

Let’s start with a standard fact, and use that as an illustration.  The city of Paris, France, is at latitude of 2.33 E.  It is convenient to use decimal notation, though the latitude is often listed as 2 degrees 20 minutes East.  Similarly, in decimal notation, the longitude of Paris is 48.8 North.  As a reference, I used an Infoplease web page.

In order to be able to express this fact, we first need the system of latitude and longitude as a way of specifying locations.  It should be obvious that the latitude and longitude system is itself a human construct, defined in effect by a system of human conventions.  The idea of using latitude and longitude was apparently first proposed by Eratosthenes in the 3rd century BC.  However its common use is more recent, due to the difficulties involved in determining longitude.

Mathematically speaking, latitude and longitude provide a system of spherical coordinates.  Such a coordinate system is standardized by specifying an axis and a fixed point not on the axis.  The latitude is then based on angular distance from the axis.  The longitude is based on the angular amount of rotation from the fixed point around the axis.  The choice of both the axis and the fixed point are arbitrary.  In practice, we use the North pole/ South pole line as the axis.  This is the most sensible pragmatic choice, as it connects the geographic coordinates with the rotation of the earth.  There isn’t any important reason to prefer one fixed point over another, but for historic reasons the Greenwich observatory is used as that fixed point.

Once we have established the latitude/ longitude system as a coordinate system, we can begin to express locations as facts within that system.  The definition of the coordinate system is itself a matter of convention.  So if we want to consider that definition to be factual, then we should consider it to consist of conventional facts (as contrasted with empirical facts).  Likewise, that the equator is at latitude zero is a conventional fact, for that latitude comes directly from the conventions.  Similarly, that Greenwich is at longitude 0, is a conventional fact.  However, once the conventions are established, then the location of Paris at 2.33E, 48.8N is then an empirical fact.  It is, however, an empirical fact that is subordinate to the conventional facts that define the coordinate system.

If we examine the physical sciences, we will find a widespread use of conventions.  These can include measuring conventions, simple naming conventions, or classification conventions such as those used by Linneaus in his systematization of biological naming.  It is likely that all scientific facts are either conventional facts or empirical facts that are subordinate to the conventions that they depend on.

When we come to ordinary non-technical language use, it becomes more difficult to track down conventions.  Where there are conventions, these are often informal conventions that develop and evolve in the cultural world of interpersonal communications.  However, it seems likely that all facts are either conventional or subordinate to conventions, where in some cases those conventions will be the informal agreements that determine the rules of the language game being followed in the natural language that is used.

Traditional epistemology and philosophy of science tend to emphasize starting with facts.  My point, in this post, was to show that they should start before there are facts.  They need to start with the establishing of conventions, upon which the facts are dependent.

This should suggest that acquiring knowledge is not so much a matter of acquiring facts, as it is a matter of acquiring the conventions which we follow in order to form facts as needed and in order to understand and use facts that are presented to us.


4 Responses to “What’s a fact, anyway?”

  1. Two questions:
    1) Is it possible for the conventions set by scientists to be non-language based, though? Your idea of ‘convention’ implicitly assumes this.

    2) Is setting up conventions very much like making up some definition of a concept? If this is the case, then we are no longer discussing how science works, since defining concepts is not an essential scientific practice. Yes, it may be helpful, but it can become quite arbitrary…so the whole idea of ‘facts’ can be tossed out as well, since it is dependent on the conventions people set up.


    • Excellent questions. Thanks. Evidently I was too vague in my post.

      1) I intended that the conventions be procedural, though typically with accompanying language so that they can be conveyed to others. Measuring conventions, classifying conventions would be good examples.

      2) I am puzzled as to why you would think that defining concepts is not an essential scientific practice. We would still be doing physics the way that Aristotle did it, were it not for the new concepts since developed. Take Newton’s laws as an example. The concept of mass was new, based on Galileo’s observations about inertia. Prior to Galileo, the nearest concept was about what we take to be weight. After Galileo and Newton, what was once the single concept of weight is now separated into two separate and distinct concepts (weight and mass).

      You are right, that concepts can be quite arbitrary. However, science constrains itself to only come up with concepts that are valuable as part of an empirical program. That’s roughly what scientists mean when ask whether a proposed theory is falsifiable.



%d bloggers like this: