On the Nature of Truth and Reality


Is there an objective nature or reality? Or is it subjective, dependent upon our perception of it? Such are the questions that fill sophomoric philosophy class essays. I will make no argument here other than to say that there is an objective reality. Anyone who says that there isn’t makes a statement that is incompatible with the world he imagines. Statements of truth belong only in a world where truth is absolute. Yet, schoolboy sophistry has little to do with the real problem. The problem is that we rarely, if ever, are able to know what reality is in an objectively verifiable way. In other words, the problem is not with truth but with knowledge.

Philosophers talk about these two issues using the terms ontology and epistemology.

Ontology, in a simple understanding, deals with the question, What is reality? What can be said to be real and what is an illusion? Is there an objective reality that exists regardless of our observation of it?

Epistemology, on the other hand, deals with the question, What constitutes belief and knowledge? How do we justify our belief or knowledge of whatever reality might exist?

In other words, ontology relates to what is and epistemology relates to how we might know it. The two issues are obviously related but they are not dependent upon one another. Whatever exists does exist regardless of our mechanisms of understanding its existence. That it, there are a certain number of grains of sand on Earth at a given point in time, even though we can’t count them; what’s more, they would still exist even if humans did not. Note that I am immediately disregarding certain extremes of solipsism which argue that the world doesn’t exist except in our minds.

Just because we have poor knowledge of truth doesn’t mean that such truth doesn’t exist, it just means that we may never know what it is for sure. It is easy to convince ourselves of truth subjectively. With too much faith in our subjective realities, and a frustrating inability to prove our subjective realities objectively to others, some have decided that truth itself is subjective (again, a form of solipsism). But this argument is the easy way out and doesn’t concern the practical world we live in.

Protagoras, the famous Greek Sophist, stated, “Man is the measure of all things,” meaning that reality was subjective, existing only as the individual views it. But Plato rejected such nonsense and stated that Sophists made their living through deception while Philosophers (like Plato) sought truth. Sophists don’t believe in truth (either in an objective reality or a personally useful concept of objective truth of such a reality).

Ironically, many who would call themselves philosophers today are nothing more than Sophists. Subjective truth doesn’t require any work to know it (it’s what’s in your head right now, after all) and sophistry is concerned mostly with making people feel better about themselves (by appealing to such narcissistic tripe) and usually making money off of them (Deepak Chopra comes to mind). But I am not concerned with such poppycock. I want to talk about how we can attempt to know the objective realities of the universe. This concept seems basic but it is of the utmost importance. It is the foundation of science and we cannot deal with any scientific issues without having a firm foundation.

What is truth?

There are many types of truths we might consider, and many famous philosophers (everyone from Plato to Kant to Bertrand Russell) have made careers for themselves (and creating a lot of confusion in the process) simply by developing taxonomies and nomenclature for types of truth. Without relabeling types of truth again, let’s think about some of these truths in order to help understand why knowledge of truth is so elusive. Let’s consider a few examples (I have no intention of being exhaustive or recreating another taxonomy).

There are historic truths. For example, there is an objective truth (ontological reality) about who killed Nicole Brown Simpson. It happened, someone stabbed her, she died, and no amount of personal bias or subjective views of ontology can change the actual events that happened (don’t over-emphasize how certain the statements I just made are – I’m speaking of ontology, not epistemology). At the same time, I don’t know for certain who killed her. I believe to a reasonable certainty that The Juice did it, but I can’t prove it (or anything else like this) to an absolute certainty. My knowledge of this truth is qualified by a degree of certainty. Even if OJ confessed to the murder, that wouldn’t matter; his confession could be a false confession. If you think about it, there is just simply no way of knowing for certain who did it. But it happened: someone did it, and our inability to be certain about it doesn’t change that objective reality.

In practical terms, though, I can show with some degree of certainty (that is probability) who killed her, and then we must decide whether the probability of the truthfulness of the assertion that OJ did it is sufficient for our purposes. This is the importance of epistemology; it needs to help us understand how we determine what is a useful, justifiable knowledge. This also means we need to define what use our knowledge should serve in order to understand what degree of probability is acceptable for justification of our knowledge.

Courts say that felony, criminal cases, like the OJ Simpson double-homicide trial, must be proven beyond a reasonable doubt; so what constitutes a reasonable doubt? That is, what degree of probability? How likely does an alternative hypothesis have to be in order to be a reasonable hypothesis? Is a 1 in 100 chance reasonable? How about a 1 in a million chance?

Civil cases demand only that a case be shown to be more likely than not – a much smaller burden. OJ was found to be the killer by a jury in a civil lawsuit. One might argue that this 51% probability standard is too low in many situations. But the point is that both standards attempt to define what degree of certainty is required of a jury; neither require that a jury be certain. More likely than not is indeed a fairly weak standard. It is more likely than not that I will survive a 45 mph car crash, yet no one will be surprised if the crash kills me. In any event, we must decide in each situation how certain our knowledge of truth needs to be in order to act.

Did the British burn the White House in 1812? I don’t know. I can read the historical accounts of many eyewitnesses and examine the evidence and conclude with a high degree of certainty that they did. But I will never know for certain if Dolly Madison didn’t do it herself because she really wanted to remodel the place. Yet for all practical purposes, I believe that the British did it. If I am wrong, it doesn’t really affect my life in any meaningful way. So the practical truth (and the probable truth) is that OJ killed Nicole and the British burned down the White House. These “facts” are part of my knowledge, and I believe both are reasonably certain (whatever that means). I believe I can justify these beliefs with available evidence.

Similarly, there are future truths. I can only state historical truths to a degree of probability. Similarly, I can state future truths to a degree of probability, though for obvious reasons the degree of probability for future truths is often much lower that for historical truths due to different types of evidence. That being said, we might be more certain about some future truths than many historical truths. For example, I believe that the sun will come up tomorrow. Every piece of information I know about this indicates that this is true and that is my knowledge of this truth. I would bet my life on it or any amount of money. But I am not certain. There are imaginable (and no doubt some unimaginable) scenarios that could occur which would prevent the sun from rising tomorrow. So I am incredibly sure but not absolutely certain that the sun will come up tomorrow.

I’m also pretty certain that I will be alive tomorrow, but far less certain about that compared to the sun rising. For practical purposes, I will live my life today assuming that both events will happen tomorrow, even though I know neither for certain and my degree of confidence about the two hypotheses is very different. Again, the point is that my degree of uncertainty has no effect on the future reality. My perception is not the reality. It is clear that there is a distinction between truth and our knowledge of it.

Actually, historical and future truths are no different, but what evidence exists and how confidently we assess that evidence varies because of the circumstances. I have more confidence that the sun will come up tomorrow than I do that OJ killed Nicole, but in both cases I have a personal belief that is sufficiently evidenced for me to go about my day. I feel confident in my knowledge of the truth, but allow for the possibility of error.

So too, current truths are transient and similarly assessed. What is current will soon be past anyway. All truths are similarly known. There really are no aspects of reality (ontology) than can be stated as knowledge (epistemology) with absolute certainty, but that does not mean that the reality or truth itself is not certain. What’s more, just because we do not know any realities with absolute certainty does not mean that reality is not knowable. Our human limitations of knowledge do not alter reality. Reality exists with or without us. Similarly, our individual experiences, biases, and perceptions do not alter reality (See Perception is Not Reality). Reality is knowable but with varying degrees of probability or certainty. Our certainty about reality or truth approaches a definable asymptote; the closer it gets, the more certain we are, approaching absolute confidence (but not certainty) in some situations.


In this schematic, a probability (P) of 1 is absolute certainty. The probability of our beliefs are represented by the curve. The curve approaches absolute certainty but remains asymptotic; nevertheless, we can for practical purposes consider the limit of our probability (0.9999… ) as certain.

At a certain level of probability j, we can consider our belief justified and the belief in turn constitutes justified knowledge. Everything with a probability less than j is opinion. The value of j varies, as described above, by its expected utility, though it is difficult to imagine anything being considered justified knowledge with a P less than 0.5. Most of what we consider knowledge is in fact opinion, and many opinions are useful and should be utilized; still, we must be careful to realize that they are still mere opinions. The confusion of opinion with knowledge plagues most fields of human inquiry.

There are many things that we state with absolute certainty which are merely constructs and are nor predictive of (external) reality. For example, we state with certainty that 2+2=4. But we are able to define this with absolute certainty only by construct. Kant called these constructions a priori forms in that they are not derived from our experiences; that is, they are not dependent on our ability to accurately understand reality. Similarly, we think of truths like my social security number or hometown as absolute and utterly knowable, but these are not certain in the same way as 2+2=4; these “facts” are subject to alternative hypotheses like psychotic delusions or typographic mistakes. They are derived from our experiences and therefore are not a priori forms. But we can say that our knowledge of these “facts” approach P = 1 well enough to consider them certain for practical purposes. Kant believed that our knowledge of reality was based on both our perceptions (which can be faulty) and a priori forms. It is important to make this distinction but also to realize that our a priori forms can serve to bias and perhaps alter our perception of reality.

Kant, like many others, failed to allow for different probabilities of certainty. He did, however, believe than an eternal reality is necessary for the creation of our subjective realities (thus rejecting solipsistic views). And this is really my point: there is an external reality and there is our personal reality; we need to understand how best to bridge the gap between these two realities and we need to understand what we practically want to do with our knowledge.

Thus, we need to define what utility we expect from knowledge of our reality. I can never know the exact value of π, but I can know it to sufficient precision to solve all practical and even theoretical problems related to it. I do not know if the chair I’m sitting in is real (maybe I’m psychotic and I’m just imagining it), but I am certain enough that I don’t plan on wasting any more time thinking that it might not be real. I’ll leave that to Deepak Chopra.

Similarly, I can never know the exact date when I first ate peanut butter, but there was a date on which I first ate peanut butter and I can estimate the probabilities of certain days and determine the most probable date. Let’s say the most probable date was March 2, 1979 (and estimate this date with a probability of 0.0015%). This type of “truth” is no different than my belief that OJ killed Nicole (which I estimate to have a probability of 99%), but the confidence is dramatically lower. This low-certainty knowledge of the date on which I first ate peanut butter is not useful to me because the probability that this knowledge represents reality is too low. It is therefore opinion at best, and should not be represented as a fact or as justified knowledge. Our knowledge of objective reality must in all cases be qualified with probability.

Our brains are wired to work this way. We gut estimate our level of certainty about anything and everything we believe. But our estimates are skewed by cognitive bias and are usually horribly wrong. It should follow that the more we work to overcome cognitive bias and the more data we have available to analyze (and the better we analyze the data), then the better decisions we should be able to make and the more probable our knowledge of reality will become.

There is an objective reality and usually we have several hypotheses for what that reality is. Thus, confronted with multiple hypotheses, we use the hypothetico-deductive model of hypothesis testing to arrange each hypothesis in order of probability (e.g., it’s more likely that the British burned the White House than did Miss Dolly). Occasionally, one hypothesis reaches a level of probability that is acceptable for our use so that we consider it as “true” and therefore it constitutes justifiable knowledge (I accept it as true that the British burned the White House). This occurs whenever P > j.

Think about this: there is a difference in a fact being uncertain and a fact being unknown. Imagine that I roll a dice; the result of the roll is not actually random. Given enough information (like the velocities and motions of the dice, wind resistance, characteristics of the surface it will land on, etc.) and the correct understanding of all the variables that affect the roll of the dice, the eventual result is actually deterministic! We will talk more about this concept later (that is, that randomness describes our knowledge, not reality; and that it is a mind projection fallacy to assign qualities that describe our knowledge onto the objective reality).

So when we throw the dice, we are uncertain how it will land; but since we can assign an a priori probability of each possible outcome, we have knowledge of how it will land – just uncertain knowledge (each of our six possibilities are equally likely unless we can find more data to help us refine our probabilities of the outcomes). How the dice will land is actually knowable (not random), we just don’t have enough to data to know it with a P > 1/6 (which is usually not enough to be considered justifiable knowledge, that is, P is not greater than j). Thus, we know how the dice will land (one of the six ways), we are just uncertain how it will land.

So we need to focus our interest on how we justify our knowledge of reality. We must ask the question, What is knowledge?