As I work up to a commentary on Jacques Ellul’s Propaganda, I realized that I would need to have a good understanding of what a fact is. One of the key points that Ellul makes is that propaganda is not simply composed of lies, it is primarily (especially in modern times) founded upon accurate details and conceptions. But is that what a fact is?
I have done some preliminary reading on this question, but the subject is too broad and deep for me to fully grasp. Rather than try to do a full review of the existing literature, then, or even a comprehensive look at existing theories, I’m just going to lay out my thoughts on what a fact is. These are informed by what I have encountered of these theories but aren’t based on any particular school of thought about fact and truth; they might be like something that exists or even identical, but that’s coincidence. This isn’t a claim at originality but at lack of reference.
In some ways, the task of philosophy is to discover rigorous explanations for “common sense” understandings of the world. This is the reason to ask what a “fact” is. If we are to say that propaganda includes facts but uses them for a biased purpose, we have to know what a fact is beyond simply “a piece of information”.
Frankly, while the question of “what is a fact” is central to this piece, the real point of this piece is to ask “how is information different than propaganda?” But that question has to arise in its proper context.
My metaphysics divides existence into four planes of experience, which are actually best imagined as concentric circles, or a series of circles inside one another. The outermost circle is totality, which is all of existence and beyond existence. It is not only all possible things but also all impossible things; it is everything conceivable and inconceivable. Even nothing is a part of totality.
Within totality is the plane of reality, which includes all things that exist physically and mentally. The real world is within reality, as are all possible worlds. The number one, though it doesn’t have any physical properties, is nevertheless real and part of reality. We’re not going to deal with “the forms” or anything similar; not to say I disagree, just that we’re going to mention that and move on.
Within reality is the plane of actuality, which includes all things which are said to “actually exist”. Primarily this means physical objects: cups, human bodies, the Earth. Could it also mean events? Could the number one fall within actuality? These are possibilities, but for the moment, we don’t need to dive deep into them. The distinction here is enough.
Within actuality is the final plane: materiality. This is the level on which molecules work, or atoms, or quarks, or whatever is inside quarks. The basic stuff of existence. We don’t perceive things in a truly material way (this is unconnected from “materialism” in a political/sociological sense), we perceive things after they have been combined in a particular material way.
Almost by definition, reality is more real than totality. That is, those elements which fall within the plane of reality have a higher degree of realness than those which fall outside of reality. Similarly, the actual (and the material) is an endo-real (or essence of real). Put another way, the real consists of the physical world (the endo-real) as well as the imaginal world (the real), with the imaginal world being built upon the physical world. There is a sense to which the material is the realest, actuality is realer, reality is real, and totality is extra-real.
Now that I have introduced this idea of levels of existence, I can give a definition of “fact”: A fact is a notion which purports to be true or false, or to have some truth value more generally. Any other definitions of fact should derive from this basic definition. However, there are two clear things to define within this definition: the first is “what is a notion”, the second is “what is truth”.
One of the failings this piece is going to have is that I am not fully conversant with philosophical jargon. If you read philosophy or any kind of social science (or science more generally, I would think, but I’m less read in the hard sciences), you’ll know that two authors are likely to use different terms to refer to the same thing, and that makes it hard to compare works. But not being fully conversant, I’m going to just have to use the words that seem best to me at the time. At a later point I will have to connect this better to the existing flow of investigation into fact. I say this because I’m now going to talk about “ideas”, “data”, etc. in ways that I know have established meanings; I would ask you to pay attention to how I am using them because it’s very likely there’s going to be some difference in how I use a term and how it is technically used in any one discourse.
A fact is a kind of notion, and a notion is a kind of idea. An idea is a concreteness which exists within abstract mental space. To define it in more detail would defeat the point. The reason for formulating “idea” in this way is similar to the reason for postulating “totality”: it establishes a complete field from which we can make further claims.
A notion is a kind of idea, like I’ve said. Specifically, a notion is an idea with a particular internal coherence. Again, a more detailed definition would actually be describing a different phenomenon, but I will try to clarify this somewhat. “Thought” (collective plural) is generally defined as “mental activity” and “a thought” (singular) as “an instance of mental activity”. With that definition, we can additionally (but not instead) say that a notion is the kind of thing which is usually able to be expressed through thoughts. A notion is not a thought, nor does an idea being inexpressible through thoughts at any moment (or even enduringly) mean that it is definitely not a notion. However, as a way to clarify the meaning of “particular internal coherence”, the condition of “being expressible through thoughts” is a quasi-standard.
I am aware that investigation into the concept of truth has been going on for thousands of years and is not settled. My definition may seem to be very simply stated and therefore not a proper reflection of the ins and outs of the debate. I can’t even properly place my idea among the major theories such as correspondence, coherence, pragmatism, etc.
Truth is a notion. It is a commonly-perceived and agreed-upon notion, in at least many respects; that is, if I say something is “true”, this statement is broadly understood by different people in a way that allows them to reach similar results. At its base, truth is the notion that some element existing in super-actual space is functionally equivalent to an element existing in the endo-real space. Put more casually, truth is the notion that there’s a connection between an idea and an actually existing thing. This thing may be an object, this thing may be a state of affairs (a term which I am using very loosely here), it may be some other thing or non-thing; the important point is that there’s a correspondence between the non-actual idea and the actual-material element.
A few things should be discussed before moving further. Firstly, in strict terms, truth is not true; the description of truth does not connect to anything existing within the actual-material world; that is to say, there is no truth-thing. In addition, this relationship does not need to be definite. To say that something is true is to say that it purports to have a relationship with something actual-material. In the final reckoning, though, there can be no certain identification between imaginal things (ideas) and actual-material things. The fact that no final truth can be proven means that no final truth can be disproven either; truth is always a subjective understanding. Following from that, something can be considered true which cannot be suggested to have a direct correlation with the actual-material.
For this to make sense, we first have to understand that every element of the world can only be comprehended in its context; that is to say, in relation with many other things. Obviously, there is an “ideal context” in which everything is understood in terms of everything else at once, but this appears to be beyond human thinking. We generally understand the world through various different contexts, some of which are complementary and others exclusive, some operating only in narrow circumstances and others working more broadly. These contexts work similarly to how grammar works for languages: they shape how an element can be understood without necessarily determining the element itself. What is most important for those who are using these contexts is that they are useful, which is to say that they produce a desired result when used. When a context or system is reliable, which is to say that it produces the expected results in a consistent manner, it is considered to be more useful than if it is unreliable; this obviously works in degrees such that a more consistent system is considered more reliable than one which is less consistent.
We have a natural assumption that we exist in a shared physical context, which we can just call the natural context, what some might call “the objective”. This is absolutely one context in which we understand the world. However, there are also different contexts, for example the context of a particular organization. It may be true that according to physical laws existing in the natural context that physical things which are released will fall. We could imagine an organization in which it would be more profitable to believe that things which are released instead go leftward, even if that is not what happened when a thing was actually released. In such a context, the notion that things fall when they are released might cause someone to believed that to be demoted, lose their job, etc. Being less useful, such a notion would be considered less true within that context (at least insofar as this organization’s context could be kept separate from the results of the actual effects).
This is an extreme conception, however. The point is to emphasize that truth is a relationship, that it is subjectively assigned (despite a common understanding), and that it merely suggests an association with materiality-actuality rather than actually having such an association. Usually, people recognize the difference between an “objective truth” (i.e. a truth in the natural context) and a “social truth” or something of the kind. My point is that the recognition of this difference doesn’t actually mean that a different operation is going on here. Both “kinds of truth” rely on the same truth idea, just in different lights.
The distinction becomes important when considering what a fact is, because the definition of a fact cannot be that it is “objectively true.” Even things which we believe to be true in the natural context are only believed to be true. It is impossible to know-with-full-knowledge that an idea has a direct correlation to an actual-material thing; this can only be assumed or believed. Therefore we cannot define a fact by an evident relationship between the fact-notion and anything actual-material.
I am aware that the definition that I’ve given so far of a fact is somewhat incomplete. I’ve narrowed down how facts are related to truth but I have not yet separated fact from opinion; if we can’t point to a final truth in order to do this, how can they be separated? For this we will need to investigate the ideas of data and information.
Here I want to offer another mea culpa, perhaps greater than before. I am going to employ the terms data/datum and information here, as well as other terms, which are related to information science. I have read almost no information science in preparation for this. The reason is that I am not actually studying information in the sense that information science does; as you’ll see, information is primarily a perspective in what I am laying out. So for those who are well-versed in information science, I am probably going to use terms incorrectly for that context; just bear with me.
The best way to understand my idea of data is to have a definition of information, at least as commonly used. Pat me on the back for getting this far without breaking out the dictionary, but I do feel it’s useful here. Three definitions from Collins English Dictionary states that information is “knowledge acquired through experience or study”, “knowledge of specific and timely events or situations; news”, and “the act of informing or the condition of being informed”. All three of these will be important for us in understanding both data and information.
We can think of data as “dis-biased information” as a starting point. If we take the first definition, “knowledge acquired through experience or study”, we can strip it of the necessary bias so that it reads something like “elements mentally perceived through experience”. We could perhaps drill down more on “experience”, but the main point here is to separate “data” from the idea of “knowledge”. In this conception, information is a kind of data or a perception of data, while knowledge is a relationship that a being has to data.
Data is a plural of datum, such that data can also be defined as “any set of things, each one of which is a datum” (or “a set of datums”); any particular group of datums is a “dataset”. There are three kinds of datum: an observation, a position (or attitude), or a non-awareness event. Perhaps simpler to understand are their expressed variations: in order, the predicate (or possibly proposition, I have to investigate this difference), the judgment (aka simple opinion), and the non-relative expression.
- A predicate is any expression that purports to having some truth value; whether or not it is actually testable, it presents itself as potentially testable. More importantly, a predicate is intended to be understood independently of the speaker; it carries its own weight. Example: “One plus one equals two.”
- A judgment is any expression that purports only to express something on behalf of the speaker. A judgment may be perceived as a predicate but, as an expression, it is understood that it is not primarily an expression of something testable but of an untestable feeling, reaction, motivation, etc.; it does not carry its own weight, the weight is carried by the speaker. Example: “One is the loneliest number.”
- A non-relative expression is any expression which is not either a predicate or a judgment. Where a predicate purports a relationship between the expression and actuality-materiality and a judgment purports a relationship between the expression and the speaker, the non-relative expression purports no relationship at all. Example: “One.”
I am using the word “purports” very intentionally here. Whether a predicate or a judgment can be judged as having a truth value by an outsider is not the issue. What separates these forms of data is something like intention. In terms of expressions, the division by intention becomes clear: a predicate tries to state something with a truth value while a judgment states something from a perspective, and a non-relative expression has no clear intention. In terms of the non-expressed versions, we can’t properly speak of intention, but some analogue still operates to separate observation from position. To adopt a semi-technical language (just for the purpose of illustration), we could call observations “first-order events”, positions “second-order events” (or “events about events”), and non-awareness events “zeroth-order events”.
A person can comprehend data; this is an essential condition. It is not always possible to comprehend a datum, but it is always possible to comprehend a dataset, even if that comprehension is that the dataset is incomprehensible. This is because a dataset is primarily understood via a set of relations, both between the elements of the dataset, between those elements and outside elements, and between the dataset as a whole and outside elements. Therefore a single datum can be non-comprehensible (in the sense of evading or resisting the ability of comprehension) while a dataset can only be incomprehensible (in the sense of being categorized into a group called “incomprehensible”). This is the difference between something “simply not making sense” and something which is “understood to be nonsense”.
This brings us to the perspectives on data. To start, we’ll dispense with “incomprehensibility” as a perspective; I bring this up to avoid having to go on an involved proof more than anything else. I’ll concede it’s possible to have such a perspective by way of not dealing with it. In terms of perspectives of comprehensible data, there are at least three basic types, with a possible fourth. These perspectives are information, opinion, propaganda, and possibly music. These aren’t exclusive, either from one another or from other possible basic types, but these are the ones I am currently aware of.
Following from the Collins definition of information, we can say that information is data perceived through its usefulness. Any dataset can be considered as information, it’s just that this information is considered “good information” if it is useful and “bad information” if it is not useful. Sometimes, a piece of data can be considered both “good information” and “bad information” depending on the person perceiving it.
The relationship between information and opinion is similar to that between predicates and judgments. Where information is judged based on its usefulness, opinion is judged based on its authenticity. An opinion can be a “real opinion” if it is considered to reflect the speaker’s consistent or avowed beliefs and behavior, and a “fake opinion” if it appears to be held only temporarily and/or for immediate convenience. Like predicates, information is expected to stand on its own while, like judgments, opinions are considered primarily in terms of the speaker.
Before moving on, I think it will be useful to explain what I mean by a notion “standing on its own”. Information is expected to be useful no matter who speaks it. To the extent that it is part of or reflective of a system/context, information has a bearing which is independent of its source. Opinion is expected to be useful in relation to the source. The usefulness of opinion is always connected to its source, such that if the source should fail in some way, the opinion would also fail.
Propaganda is where the correspondence ends, as it is not correlated to non-relative expression. Instead, propaganda is a kind of reverse view of information. If information is meant to be useful to direct intentions within a certain context, then propaganda’s purpose is to change the context such that certain intentions are more or less viable. We could view propaganda as simply a species of information but, for the moment, we’ll leave it as a separate basic type.
If there is a correlation with non-relative expression in this set of things it is with music. What is the difference, after all, between human song, birdsong, and brooksong? Only how we receive the sounds. That is to say, there is no inherent relation between music and other forms of understanding. This is not to say that music cannot be composed with intention, or that music can’t include words, or use musical symbols; all of those things are true and possible. Musicality also cannot be reduced to patternizing; though we generally feel patterns to be more musical, it is not simply that a thing has a pattern that makes it musical, and it is not the case that not having a pattern makes a thing unmusical. Music (or musicality) can only be described, in an essential sense, as a perspective on data which is not information, opinion, or propaganda, but which is nevertheless comprehensible.
The question at this point, for those who recall where we started is this: “are facts information?” or, in a different way, “is a condition of information that it is factual?” The answer is, plainly, no. This is for two reasons. Firstly, facts play a role in information, opinion, and propaganda. The same fact may be considered information, opinion, or propaganda depending on who is perceiving it. Therefore we can’t use factuality as a criteria for something being information. Secondly, information-as-data can be divided into two categories: facts-as-observations and findings-as-positions. Non-awareness events (and their correlates) can generally be ignored as information, usually being non-comprehensible.
In terms of a dataset (in whichever perspective), the primary quality of a fact is not its truth value but whether or not it stands on its own as described above. Therefore it is possible for a “false fact” to exist; in this case, “false” can be replaced with “not-useful”. The primary quality of a finding is that it does not stand on its own but rather is dependent upon some being which makes the finding: the finder.
This raises a new question: how can a finding be information? Or rather, how can a finding be good information? Or, to be more pointed, why would a finding be used instead of a fact? “Instead of” is important here, because we can identify an obvious situation where a finding would be used “rather than” a fact: when a fact is not available. But this circumstance does not show that a finding can be information: why would we not simply say that findings are always opinion and never rise to the level of information? The essential reason that findings have to be of an equal potential weight with facts is that the context a datum is used in ultimately determines its usefulness. The easiest way to argue this point is to note that most of what we understand as facts are not facts-as-such but are instead findings of fact.
To know a fact, a person must directly experience it in some way. For me to comprehend the fact that a cup is blue I have to be able to see it and connect it with the idea of blue. Someone who is unable to see the cup may be able to know that the cup is blue in a way that is functionally equivalent to being able to comprehend the fact of the cup’s blueness personally, but they can only do this if they are related this fact from someone who can see the cup. If they’re relayed the data about the cup’s blueness, they do come to know about that quality, and therefore this data can be considered information. However, we can’t say that what the person who doesn’t see the cup is comprehending a fact in the same way that the person who sees the cup is. Instead, the person who can’t see the cup is relying upon a finding of fact. While the person who relays the cup’s blueness is themselves comprehending that blueness, what they relay to the other person is not the fact itself but the notion that the observer has seen the cup’s blueness. Each part of this message is important to the outside person: there is the “fact itself” being communicated and there is also the activity of observation. It isn’t simply that the outsider understands that the observer has seen a blue cup, it is also that the outsider understand that the observer is in a position to be accurate because they know that the observer is in the presence of the cup. In other words, when considering a finding of fact, it is not merely “the fact itself” which matters, as it would be for the observer, but also the manner of the fact’s transmission.
There are many things to consider when trying to qualify the transmitter or source of a finding. There is past reliability, perception of closeness, general competency, delegated responsibility, etc. These all deserve investigation. If we are to give this a general quality which encompasses all of these, we could call it “relevant ascendancy” or “authority”. When a source has a high degree of relevant ascendancy for a context or situation, their findings are considered to be more useful, and vice versa. This consideration of usefulness happens before the finding is used, not during; that is to say, the fact that a finding fails in its use has no bearing on whether the finding itself is considered useful; instead, it damages the source’s relevant ascendancy. An investigation of how relevant ascendancy operates, and how influenced it is by common perception, is better handled at another time. For the moment, it is enough to understand that a finding is considered to be more informative as its source is considered to have a higher relevant ascendancy. This ascendancy is relevant to the context in which the finding is being used; in other words, this is not a general ascendancy but a particular one. In terms of a scientific breakthrough, for instance, a scientific institution would have a higher relevant ascendancy than a state government, even though the government would have a higher relevant ascendancy in terms of foreign policy and other matters.
While I say that “authority” can be used in place of “relevant ascendancy”, this is only a thematic correspondence. That is, in a place where you might say “this institution has authority” you could instead say “this institution has relevant ascendancy”. In use, though, “relevant ascendancy” actually maps better onto a term such as “expertise”. Another inelegant description would be “perception of possession of useful associations”. The point I’m getting at is that “authority” as in relevant ascendancy is not equivalent in any way to “domination”, as it might be in other uses.
As should be obvious, the utility of findings is that they allow things to be known which an individual has not experienced. This reduces the effort needed for an individual to gain that knowledge. It also allows individuals to review their own experiences, as it is not true that experience is always comprehended as a true fact (by accident or incompetence), and it may be true that a finding is more useful than one’s own experience.
Have gone through these deductions, I’ve come to a workable definition of a fact: a comprehensible datum, or unit of data, which purports the ability to be evaluated independently according to a truth value.
Thank you.
Works read:
- Propaganda by Jacques Ellul
- How Propaganda Works by Jason Stanley
- Manufacturing Consent by Edward Herman and Noam Chomsky
- “Delegated Expertise” by Joel Demski and David Sappington, Journal of Accounting Research, Spring 1987, Vol. 25, No. 1 (Spring 1987), pp. 68-89
- A Theory of Semiotics by Umberto Eco
- Propaganda, Communication and Public Opinion by Bruce Smith, Harold Lasswell, and Ralph Casey
- “Facts” from the Stanford Encyclopedia of Philosophy
- “Truth”, “Fact”, “Information”, “Truth-bearer”, and others from Wikipedia