As a psychometrician, I would think the case here is more complicated than that. Disclaimer: I'm not an expert on reading speed, so I might be wrong about some things here.
I agree with you that there are probably strong correlations between the speed at which different people can read different texts, to the point where it seems like it would make sense to rank people by reading speed in a fairly unidimensional way. However, I would also agree with him that the speed at which people read things depends on other factors, such as the kind of material one is reading, or the way one is reading it.
This becomes a problem when one has to quantify the reading speed. You can't exactly quantify reading speed as e.g. "words per minute", because the number will depend on other factors. And I bet there's no known natural unit to use for reading speed.
In psychometrics, you usually solve this by just working with relative rankings of people, rather than independently meaningful numbers. So for instance if there was two people whose reading speed you both know very precisely, where one of them is a very quick reader and one of them is a very slow reader, you could use proximity in reading speed to those two people to discuss it numerically. However, I think it is quite rare for people to have clear baselines like this? Idk.
I mostly agree with the actual point being made here: a thing can be real even if the simplest possible way of operationalizing it turns out not to match reality. But, on the specific matter of reading speed ...
I (like a lot of people around here, I suspect) read very fast. But there's a problem with this. When I was learning to read and forming habits around how I do it, the material I was reading was all pretty easy. These days, I often want to read things like mathematics papers, physics textbooks, and the like, which I cannot absorb effectively if I am zipping through them at high speed -- but zipping through them at high speed is what my body and brain naturally try to do. My eyes move across the page at a rate that makes sense if I'm reading a not-too-literary novel just for fun, and if I'm reading something difficult and trying to understand it deeply this is not helpful.
If someone else is about as smart as me but for whatever reason didn't learn the habit of reading very fast, then they might actually read technical stuff faster than me because they don't keep finding themselves skimming over its surface and taking very little in.
If someone else is about as smart as me, did learn the habit of reading fast, but is better than me at adapting the physical actions of reading to different material, they too might read technical stuff faster than me.
Call me and these two hypothetical people A,B,C in order. Then when reading easy stuff we'll have A=C>B, and when reading difficult stuff we'll have A<B=C. So different people are in different orders depending on what material they're reading.
(Of course this may happen in more straightforward ways -- e.g., everyone will make faster progress on material they understand well, so if I know more about X and less about Y than you do we may have reversed reading speeds on those two topics. But the mechanism above is more interesting :-).)
So perhaps the right way to look at this is: "exists" is too binary. Reading speed is less real than height, but more real than (say) ability to perceive demonic possession (assuming, as I would expect most readers to agree, that demonic possession is not in fact a real thing). In at least some cases (of which reading speed is one) you can get something "more real" by nailing down details, at the cost of missing some of what the concept is meant to be about.
I disagree with you, and also with your friend. Reading speed does not exist in a vacuum, and it's absolutely correct to push for more specific context if you ask the question. It's also NOT correct (probably; it's reasonable to put quite minimal effort into concepts that might be useful in some cases, but not ones you currently care about) to reject the concept outright. Whether it's correct or not, it's rude to reject it without some exploration of what (if any) value it has. It's also rude to try to convince someone if they've flatly rejected it - sometimes you really don't have the bandwidth to resolve such differences of opinion.
"how quickly can you read this sample text thoroughly enough to answer 10 questions afterward" is a pretty objective question. The debate can then turn to whether it's useful for anything, and what you can predict/improve outside of that test based on the results.
It seems like I'm pretty bad at communicating what I wanted to communicate. Multiple people in the comments said, is that the case is more complicated, then assigning a single number to reading speed. I agree with this. I think that the way reading speed is normally used by people, is probably too simplistic.
My criticism is about, dismissing the concept as unworkable. There are various algorithms and data in the brain, that enables a human to read. Depending on what this algorithms do exactly and on what kind of data you have, you will be in some sense worse or better at reading. There might be many dimensions to what it means to be good. But there are underlined data and algorithms that in principle we could understand. And I think we could understand them so well, that we would not be confused anymore at all, about what definitions of reading speed make sense for which purpose.
Dismissing to concept of reading speed, in some sense also dismisses this possibility. For rather after you dismiss the concert, would not think about in these terms. You would not think about this at all anymore, as you would have successfully handicaped yourself, by removing a concept that captures something about reality.
This is a general thing that I have observed people doing. And I think doing this is always bad, as long as the underlying concept actually talks about something in the real world.
I think, in order to defend "reading speed" as a useful atomic concept (or perhaps a cluster of things with this as a proxy measure), you would need to define more closely exactly what you think it is, not just waving toward "something in the real world".
I don't fully disagree with you that improving one's speed of absorbing information is useful. I've studied and practiced speed reading, including time trials and tests of words-per-minute with retention quizzes. I recommend doing so to almost anyone reading this (indicating that you read things regularly). Even so, I don't think it's a "real thing", but a set of capabilities that are imperfectly measurable; "reading speed" is a correlate of some real things, not a real thing itself.
I don't mean to offend, it might be my fault, but I don't think you got the core idea that I was trying to communicate. Probably because I did not say it clearly, or maybe it is to be expected that some people will always not get the core point? But that sounds like an excuse. My core point is not that reading speed is a good thing to improve on (though it might be). It is merely an illustrative example. An example that is supposed to illustrate the core thing that I am talking about, such as to make the general abstract pattern that I want to convey more concrete.
The general pattern is that there is a concept C that is very vague, or even flawed in important ways, but nonetheless points at something in the world, that seems important to have a concept for. Then somebody comes along and says "C is flawed in X way, therefore we should not even use it" or something like that. My point is that abandoning a concept like this, which actually captures something true about the world is almost always a bad idea if you don't have another way to capture the true kernel that was captured by the original flawed concept.
Instead, you should be aware of the flaws of the concept and use it appropriately. Trying to fix it can be very good. But just abandoning it is almost always dumb IMO.
I feel like he was falling into a kind of fallacy. He observed that a concept isn't entirely coherent, rejected the concept.
My go-to writeup on this is Luke Muehlhauser's Imprecise definitions can still be useful section of his What is Intelligence? MIRI essay written in 2013, which discusses the question of operationalizing the concept of "self-driving car":
...consider the concept of a “self-driving car,” which has been given a variety of vague definitions since the 1930s. Would a car guided by a buried cable qualify? What about a modified 1955 Studebaker that could use sound waves to detect obstacles and automatically engage the brakes if necessary, but could only steer “on its own” if each turn was preprogrammed? Does that count as a “self-driving car”?
What about the “VaMoRs” of the 1980s that could avoid obstacles and steer around turns using computer vision, but weren’t advanced enough to be ready for public roads? How about the 1995 Navlab car that drove across the USA and was fully autonomous for 98.2% of the trip, or the robotic cars which finished the 132-mile off-road course of the 2005 DARPA Grand Challenge, supplied only with the GPS coordinates of the route? What about the winning cars of the 2007 DARPA Grand Challenge, which finished an urban race while obeying all traffic laws and avoiding collisions with other cars? Does Google’s driverless car qualify, given that it has logged more than 500,000 autonomous miles without a single accident under computer control, but still struggles with difficult merges and snow-covered roads?4
Our lack of a precise definition for “self-driving car” doesn’t seem to have hindered progress on self-driving cars very much.5 And I’m glad we didn’t wait to seriously discuss self-driving cars until we had a precise definition for the term.
Bertrand Russell put it more pithily:
[You cannot] start with anything precise. You have to achieve such precision… as you go along.
I basically agree with this. But if you apply what are described in the post, it's reveals a lot about why we are not there yet. If you pit a human driver against any of the described autonomous cars, they will just be lots of situations, where the human performs better. And I don't need to run this experiment, in order to cash out its implications. I think when people talk about fully autonomous cars, then they have implicitly something in mind where the autonomous cars at least as good as human. Thinking about an experiment, that you could run here, makes this implicit assumption explicit. Which is think can be useful. It's one of the tools that you can use to make you definition more precise along the way.
The problem with the concept of reading speed is that focusing on it as a scalar metric led to speed reading literature that recommends a bunch of habits that seem to reduce text comprehension. Optimizing for reading as fast as possible while having a sense of subjective comprehensibility doesn't seem useful.
One corollary of the GPT-based language models is that it should now be easier to create an app that actually trains reading comprehension by autogenerating questions after reading a text.
Today, I talked with somebody about reading speed. I asked him how fast he can read. He didn't answer, instead he said that the concept is abused by people. He said, it's more complicated than to say that you read at a certain reading speed. It depends on if you're reading a novel, a history textbook, or a poem.
I feel like he was falling into a kind of fallacy. He observed that a concept isn't entirely coherent, rejected the concept. However, the concept of reading speed seems real. It seems to capture something about reality.
This becomes obvious once you think about an experiment where we have two people that read the same material and time them. I read "The adventures of the Lightcone team" (or whatever it is called) together with Chu. We made the bed that I can get more than halfway through the book, before she finishes. I bet $5 on that. When she was finished, I almost managed to get halfway through the book. I was trying to read really fast, at the edge of comprehensibility.
Clearly there are latent causes, in each of our brains, that determine how fast we can read while still comprehending the text.
Trying to operationalize the concept that you're talking about, and imagining what sorts of experiments you would try to run to measure it, might be a good general way to avoid the fallacy of dropping a concept and losing its true kernel. Often you don't even need to run the experiment. Imagining it is sufficient.
Edit: see also this follow up comment.