What is Critical Thinking?
Is there such a thing as critical thinking? Some people seem to think that there isn’t. Check this post on X:
Greg Ashman draws here on a study claiming that general thinking skills are nothing more than disguised expertise in particular domains, where the knowledge and skills developed by focusing on specific tasks are mistaken for some wider thinking ability.
I am a philosopher, and a professor of critical thinking and this problem strikes at the heart of my professional identity. Is my job of teaching students how to think critically doomed? Am I wasting my time?
Perhaps they are right. Sometimes even I have doubts.
But, maybe the problem is in what we think critical thinking is. Critical thinking is notoriously hard to define because it refers to not just one, but several skills.
For example, according to Edward M. Glasser, critical thinking is:
the intellectually disciplined process of actively and skillfully conceptualizing, applying, analyzing, synthesizing, and/or evaluating information gathered from, or generated by, observation, experience, reflection, reasoning, or communication, as a guide to belief and action.
Similarly, folks at Monash University in Australia define it as:
a kind of thinking in which you question, analyse, interpret, evaluate and make a judgement about what you read, hear, say, or write.
Per these definitions, to be a good critical thinker you need to be good at the following:
asking questions,
analyzing and interpreting data, and
making value judgments
Some people think that there is a common thread behind these. Stanford Encyclopedia of Philosophy (SEP) says they all build around the same core of ‘careful goal-directed thinking’.
That seems true enough, but if you want to become a critical thinker you need to master all of those skills. There is no single (and simple) recipe for developing critical thought.
So, could it be that critical thinking is just terribly complex, and very hard to learn on its own?
Critical Thinking and Education
Critical thinking is not just an intellectual process or a disposition but also an educational program. Most universities worldwide include courses on critical thinking. Many serve as prerequisites for further academic progress; students cannot take other courses unless they have passed their Critical Thinking 100.
So, how are these skills taught to students?
While approaches from school to school (and from instructor to instructor) vary, the most common pedagogical structure includes modules on:
Deductive reasoning (basics of logic), and
Inductive reasoning (basics of statistical thinking).
Additionally, these are expanded by further emphasis on logical fallacies, inference to the best explanation, and other corresponding topics. This is how I have taught critical thinking for years.
Many professors also choose to focus on more interpretative skills, where students read philosophy classics, such as Plato (or some of his contemporary disciples), and learn to develop good habits of thought by observing how the OG’s do it.
While there is (in principle) nothing wrong with this approach, I think we can improve it.
Why does it need improvement? For a simple reason:
So far nobody (that I am aware of) offered a stronger unifying narrative that can bring these distinct skills in closer connection to one another. Sure, SEP’s ‘careful goal-oriented thinking’ provides an accurate description of it, but it lacks motivational content. There is no ‘punch’ in this narrative frame; it is too bland and reminds one too much of corporate training lingo.
As philosophers (especially those of non-analytic bent) know, critical thinking has to integrate human existence into the picture. It has to address passion, beauty, anxiety, and all other elements of the existential dread (or pleasure) we feel every day.
Otherwise, students are getting a very unrealistic picture of what it means to be a critical thinker in the real world, not just in the classroom. Consequently, they find it very hard to apply the classroom lessons in their real life, and we find it hard to recognize when students have mastered these skills (and even to accept that critical thinking is a skill itself).
So, how would a stronger unifying narrative help? What would it be based on, if not ‘careful goal-oriented thinking’?
Entropy
The concept of entropy is challenging. It was first introduced in the mid-19th Century and since then it has swept through many scientific disciplines, from thermodynamics and statistical mechanics to information theory and cosmology.
Yet despite the complexity brought by interdisciplinarity, the core of the concept is relatively simple: in essence, entropy is understood as a measure of disorder.
Even if you’ve never thought about how to measure order and disorder, I’m sure you can easily tell the difference between the two: at any point in time, you can tell whether your room is messy or tidy. Are the books neatly stacked on shelves? Are your clothes neatly folded in the closet? Dishes washed and put away?
When your room is messy, it is in a higher state of entropy. When it is tidy, entropy is low.
I think entropy, broadly understood, is a perfect candidate for a unifying concept that can tighten up our narratives about critical thinking, and help us think better.
Here’s how it can do that.
It Freaks Us Out
First, entropy is not just any measure. According to the Second Law of Thermodynamics, a cornerstone of our understanding of reality, total entropy (in a closed system) tends to increase over time.
This means that for us things naturally tend to disorder.
In essence, things slowly fall apart and become chaotic and disordered. It is an irreversible process, akin to time itself.
The most obvious way to notice this is the simple fact that we’re slowly dying. Our bodies slowly deteriorate and end up scattered as particles in the universe, assuming a much higher state of entropy.
This is scary for us. Death is not a desirable state; we avoid even thinking about it.
Yet, it is pervasive and shapes the way we think and behave. There are so many actions we do out of simple fear of death, often being completely in the dark about the root motivations.
Here are a few examples:
Seeking fame and legacy: We often strive for fame or to leave a legacy as a way to achieve a sense of immortality, hoping to be remembered after we’re gone;
Accumulating wealth: We pile up possessions as a way to feel more secure and in control;
Obsession with beauty: We develop an obsessive desire to prolong aging, and resort to all kinds of interventions to hide it (plastic surgeries, etc).
To be good thinkers, we must be aware of the forces that shape our actions. Without a unifying narrative explaining why death occurs and how it shapes our thinking, it is hard to shine a light on thinking practices that can counter the fear of death. Very little in the existing critical thinking curricula does this.
With entropy in the center focus, that would change. Students can learn that thinking is not just an instrumental activity to reach a certain goal, but an existential one. Critical thinking is our way to grapple with our most terrible fears.
‘I think, therefore I am’ is not a motto of a ‘goal-oriented thinker’: it is the cry of a creature faced with the certainty of death.
Thinking as Ordering
Second, we can understand thinking as a process of ordering the messy everyday experience. Inputs to our senses are nothing more than raw data we have to make sense of.
Fortunately, our brains have evolved to do just that: they are good at picking up patterns that help us understand the world and thrive in it.
So, reducing entropy is natural to us.
However, not every ordering system is equally good. We are used to it so much that we sometimes create order when there is none, or create orderings that do not represent reality. This is how superstitions are born. Blowing the dice before rolling them is an example of such an order: it is a belief that when rolling the dice is preceded by some action, the outcome will be favorable to us.
The concept of entropy can help here by showing that order is not total, but continuous. Very few things are absolutely ordered or disordered; most of them are somewhere on the continuous line.
Even math, our prized pinnacle of organization, escapes complete ordering.
Entropy is closely related to the notions of randomness, and probability. It can help us understand the difference between order and disorder, even when we’re dealing with unpredictable outcomes.
Here’s an example. Imagine we’re tossing two fair coins ten times. Here are the outcomes of those two tosses:
Coin 1 = [Heads, Heads, Tails, Tails, Tails, Heads, Tails, Tails, Heads, Tails]
Coin 2 = [Heads, Tails, Heads, Tails, Heads, Tails, Heads, Tails, Heads, Tails]
What do you notice?
Imagine we toss the same two coins a couple of hundred times, and the outcomes of both coins look similar to the ten-toss example (with no particular order in the sequence of tosses in Coin 1 and the alternate ‘Heads, Tails’ pattern in Coin 2).
If the coins were fair, then the probability for either ‘Heads’ or ‘Tails’ on each toss would be 1/2 and you would not be able to predict the outcome of the next toss with much confidence. According to probability theory, any sequence of heads or tails would be equally probable to any other (since each toss is independent).
However, the fact that the outcome of Coin 2 is patterned would give you enough reason to believe that something is ‘fishy’ with this coin. The coin would be predictable because of its inherent order and low state of entropy.
Critical thinking, although complex and multifaceted, revolves around the process of reducing entropy imposed by our sense experience. It is domain-independent to the degree that it helps us create successful and accurate models of the world (orderings that work).
Why Thinking is Hard
Entropy can also help us understand why critical thinking is hard.
As any student who prepares for an exam knows, thinking is tiring and requires a lot of energy. Our brains are the most energy-consuming organs in our bodies (this could explain why students devour everything in their path during exam week).
Thermodynamics teaches us that the only way to reverse the process of ever-increasing entropy is to expend energy. In other words, reversing the path from order to disorder requires effort.
This is a basic fact about our universe. There is no escape from it.
We try to minimize thinking effort by relying on shortcuts. For example, we decide to trust some sources in advance, so we don’t check the veracity of their claims every time (do you double-check things you read on Wikipedia?).
Cognitive psychology calls these shortcuts heuristics, and they are often the source of biases and fallacies that prevent us from thinking critically.
Here’s a few of them:
Anchoring Bias: We often rely heavily on the first piece of information we hear (the "anchor") when making decisions. For example, if the first car you see at a dealership is priced at $30,000, you might base your judgment of all subsequent prices around this figure, regardless of the actual value of the cars;
Confirmation Bias: We usually favor information that confirms our existing beliefs or hypotheses. For example, if you believe that left-handed people are more creative, you are more likely to notice and remember instances that support this belief and ignore instances that don't;
Bandwagon Effect: We all have an embarrassing tendency to do (or believe) things because many other people do (or believe) the same. For example, buying a product because it’s popular, assuming it’s good because everyone else is using it.
Succumbing to biases and fallacies like these happens because we are often too lazy (or tired) to think hard enough. The problem is exacerbated by our lack of understanding of the relationship between critical thinking and energy use. Understanding thinking through the lens of entropy could help. If critical thinking is ordering, and if ordering requires effort, then any true critical thinking must be hard and energy-consuming.
Critical as Computational Thinking
I hope I persuaded you, at least a little bit, that the concept of entropy can serve as a natural glue for disparate skills that comprise the elusive ability we call critical thinking.
But, if we want to translate this connection to something pedagogically useful, we must provide more structure. What we need, for lack of a better word, is a critical thinking syllabus that can connect entropy to concrete reasoning skills.
In this section, I will describe what I think is a good intermediate method of understanding critical thinking as an entropy-reducing educational program.
Since the concept of entropy seems more at home in the technical disciplines, rather than in humanities, I think the best possible framework is offered by the set of ideas gathered under the umbrella of computational thinking. Given entropy’s existential dimensions, I believe it is also perfectly poised to bridge the gap between humanities and sciences and present a unified front for the advancement of thinking skills.
Computational thinking refers to a problem-solving approach centered around four basic practices:
Decomposition,
Abstraction,
Pattern recognition, and
Algorithmic design.
It is understood as a conceptual framework for the education of computer scientists, but it has recently received (well-deserved) attention from educators from other disciplines for its useful pedagogical properties.
Here’s how these practices relate to entropy-focused critical thinking:
Decomposition
One of the most important mental dispositions responsible for good practices of thought is clarity. Clear thinking is hard to achieve, yet it is a prerequisite for understanding and communicating ideas.
The best way to achieve clarity is to practice decomposition. This refers to the process of breaking down ideas and thoughts into their basic elements.
But, what are the basic elements?
We can use logic to figure that out.
Thoughts and ideas are expressed using language, and every language follows some basic logical structure. Take propositional logic, for example, which uncovers logical relationships and properties of propositions by decomposing them into constitutive elements.
Propositions are sentences that can be either true or false (they have ‘truth value’). They are also sometimes called ‘statements’ (because they state that something is true).
Here’s an example:
p = ‘It is raining.’
q = ‘There are clouds.’
These two are the most basic propositions we can have. Any sentence shorter than this would not convey any meaning. Propositional logic allows us to assign these to some variables (I chose p and q here) and use them to make further combinations that convey more complex meanings. For example,
p AND q
means ‘It is raining and there are clouds’. Similarly,
p OR q
means, ‘Either it is raining or there are clouds.’ And,
IF p THEN q
means, ‘If it is raining, then there are clouds.’
You get the idea. Propositional logic helps us decompose thoughts expressed through language down to their most basic elements so we can see more clearly what these thoughts exactly are and how they relate to one another (if you’re a computer scientist you’ll recognize the similarity with Boolean logic).
This is particularly useful when we get to more complicated propositions that we struggle to understand. Whenever we are faced with some cryptic piece of writing whose meaning is uncertain to us (it is in a high state of entropy), we can use decomposition to achieve order and understanding or to conclude that some piece of writing lacks meaning or truth.
Abstraction
Once we learn how to decompose thoughts into basic elements, we can abstract away from particular propositions and focus exclusively on the structures that underlie complex propositional constructs.
For example, we can learn that different types of logic can help us represent different types of thought. Besides propositional logic, we can use other types, such as predicate logic (which adds quantification to our propositions, so we’re able to convey more precise meanings about parts or wholes in our thoughts), or modal logic (which builds on both propositional and predicate logic and adds notions of possibility and necessity).
Each foray into abstraction from our daily experience mediated by our senses and our language helps us explore possible ways to bring order to different types of experiences and do two crucially important things:
Learn how to make sense of our experiences, and
Build on them to develop tools to improve our lives.
It is important to understand that abstraction is not the main goal here. We don’t abstract for abstraction’s sake. The point of any abstraction is to get back to the real world and solve some concrete problems.
Every piece of technology that has ever been invented (from a walking cane to an iPhone) was a result of a process of abstraction, where abstraction was just an instrument of problem-solving.
Pattern Recognition
Once we can decompose thoughts, claims, and arguments into their constitutive elements, we can recognize patterns, and start distinguishing between successful and unsuccessful ones.
This allows us to see that, for example, when we make an argument following this pattern:
IF p THEN q
p
Therefore, q
the truth of q will be guaranteed, but when we do the opposite:
IF p THEN q
q
Therefore, p
the truth of p will not (it may be more likely, though).
The difference between the two can be explained in terms of entropy. The first example (the famous Modus Ponens, or ‘Affirming the Antecedent’ pattern) has zero entropy because of its absolute order. If the first two propositions are true, the third one must be true by necessity. There is no uncertainty in our minds about the truth of q in the conclusion.
In contrast, the second one has much greater entropy because p is merely probable, and not certain (it is only partially ordered). To see this more clearly, let’s plug in real content to these propositional variables:
IF it rains THEN there are clouds.
There are clouds.
Therefore, it rains.
Although rain may be more likely on a cloudy day, it is not guaranteed. While examples like this are considered invalid in classical logic, we could still admit that they could be useful to a certain degree due to their partial order. Entropy helps us see that.
Algorithmic Design
Finally, once we get better at decomposing, abstracting, and recognizing patterns, we can start being creative in how we process our thoughts and move forward.
Algorithmic design is a bona fide critical thinking skill, and it should not be reserved only for computer scientists. It helps us ‘connect the dots’ and make leaps in understanding, reasoning, and problem-solving in any domain of life, not just programming.
This is where critical thinking gets to be creative in ways not unlike art. Any act of creation is an act of creating order from chaos: painters combine colors to create an ordered representation of their experience; composers combine sound frequencies in orders that please our eardrums; and poets order words in ways that elicit strong emotional reactions in us. They all ‘fight’ entropy through their work.
Algorithms are usually understood as sequences of steps that achieve a certain goal. Although true, I think this definition minimizes the importance of algorithmic design for overall thinking skills.
Humans are goal-oriented creatures, and whatever we do, we do it with a certain goal in mind. For any possible goal, there is (usually) more than one ordered way of achieving it. You can elicit an emotional (re)action in a person by pointing a gun to their head, or by writing a poem. Which one of these is better (by whichever measure you use) depends on your ultimate goals, constraints, and costs. A creative way to problem-solve will optimize your chances of achieving the goal within constraints and minimal costs.
Another useful aspect of algorithmic design is that it can help us understand complexity in a much deeper way. We cannot achieve any complex goal through a simple algorithm. Anything worth doing requires not just a lot of effort, but also creative problem-solving that includes many intricate steps that may not be easily taken. There is an inherent (and necessary) connection between the complexity of phenomena in the real world (whether they are experiences, artifacts, or creations) and the complexity of algorithms that could produce them. Learning about this connection can help us understand and appreciate order when we see it.
Conclusion
I had to stop writing at some point, so I decided to do it here.
What I wanted to say with this article deserves more space (and time). I will continue to explore each of the subtopics in separate articles, add more details, and offer more arguments in support of a claim that critical thinking can be taught as a standalone set of transferable and domain-independent skills built around the concept of entropy.
Until next time, thanks for reading!
Feel free to leave a comment, suggest an idea, or simply share your thoughts.
Can't say how much I enjoyed every part of this. I've been planning for a long time to write some sort of short free ebook precisely on the topic of computational thinking, for college students of all subject matters. A kind of manual on how to develop those critical thinking skills taking what's best in the computational disciplines. What do you say? Should we do it?
(Btw I think you have a couple of sections repeated, from abstraction onwards, maybe it was my reader)
I'll dump a few random thoughts I had while reading here, in no particular order. I'm not sure whether some of these observations are "obvious" to others or not. They may be.
- I think a lot about teaching methods, which is understandable since it's what I do for a living. I've been doing quite a bit of additional thinking recently as I tried to identify the components in my teaching that matter most to me and benefit my student mostly. Why am I saying this? Because as I read the definitions you quoted of critical thinking, I couldn't help noting that they applied perfectly, almost word for word, to good teaching:
"the intellectually disciplined process of actively and skillfully conceptualizing, applying, analyzing, synthesizing, and/or evaluating information gathered from, or generated by, observation, experience, reflection, reasoning, or communication, as a guide to belief and action." Just alter the last few words, and this is a definition of teaching
- This is probably not that surprising, after all. Perhaps the skills needed to be good at critical thinking overlap significantly with the skills needed to be a good educator? Perhaps they're necessary skills (but not sufficient)
- In my effort to decide what's the most important aspect of my teaching, "clarity" is one of the concepts I ranked very highly–I've been using the term a lot more, and I think I included it in my new LinkedIn profile bio and X pinned post. So when you talked about the importance of clarity in the decomposition process, that again resonated with me as essential to good teaching, or rather, I shouldn't use "good", so I'll say "my preferred method" of teaching.
You claim that "Clear thinking is hard to achieve, yet it is a prerequisite for understanding and communicating ideas." The teaching process is trying to make this process easier for learners through the way educators communicate. And, as we often hear quoted, the best way to learn something well is to teach it, since you need to have that very clear perspective of the topic before you can teach it well.
- So, our job as educators is: first to reduce the entropy surrounding a certain topic as much as possible for ourselves, then to find ways of helping our students reduce their own entropy. Hope I'm using the entropy analogy well here! I skirted around this idea a bit, not using entropy but another concept from physics, in one of my Breaking the Rules (my _other_ Substack) articles: https://breakingtherules.substack.com/p/a-near-perfect-picture-ep-7
- Decomposition, Abstraction, Pattern recognition, and Algorithmic design: you already mention this yourself. This is a great framework for teaching any subject. I won't dwell on this further in these comments.
I'm looking forward to reading more in your next articles, and thinking (hopefully critically) more as a result…