Meister and Zheng's Nonsense Claim That Humans Live at Ten Bits Per Second

Ah, the cushy life of the modern biologist. He can do a poorly designed experimental study, and his bungling research can get published (being approved by peer-reviewers who often do research just as bad). His bungling research can get trumpeted by the uncritical "echo chamber" guys that call themselves science journalists, but who typically act more like cheerleaders rather than good journalists. Or he can write some study making the most glaring factual errors and the laziest type of scholarship, and if he has produced something interesting-sounding, a horde of websites eager for profitable clickbait will make sure his claims are repeated on a hundred web pages. To get the hype ball rolling, it helps to have a typical university press office, populated by writers willing to engage in the most shameless exaggeration, trying to make dross results look like gold. 

A recent example is the bungling paper "The Unbearable Slowness of Being: Why do we live at 10 bits/s?" by Jieyu Zheng and Markus Meister. The paper makes the extremely untrue claim that the "throughput of human behavior" is  10 bits per second. The authors state, "the information throughput of human behavior is about 10 bits/s." Throughput is defined as "the amount of material or items passing through a system or process."

The authors quickly tell how they got this idea. They say that good typists typically type at 120 words per minutes. Erring by a factor of 1000%, the authors then give some utterly erroneous calculation trying to suggest this is a speed of only 10 bits per second. We read this:

The reference is to some old paper behind a paywall. Whatever obscure thing "entropy per character" is, it sure isn't the number of bits per character needed to transmit or store a stream of typed characters. Doing a Google search for "number of bits in one character" will give you an answer of "8 bits."

Our authors have gone way, way wrong. The information content of written English is a fact that almost everyone agrees on. An English character contains about one byte of information, as innumerable writers have stated. One byte is 8 bits of information. The average word in English has about five characters, and requires about 40 bits of information to store or transmit. Since people need to type an average of one space character per word, it effectively requires 48 bits of information per word typed. A person typing at 120 words per minute is producing output at a rate of about 48 times 120 bits per minute, which is 5760 bits per minute, which is a rate of 96 bits per second. By claiming that someone typing at 120 words per minute is producing output at a rate of only 10 bits per second,  Zheng and Meister have produced a calculation that is a wrong by a factor of about 960%. 

If you doubt my claims above, consider how an English character is stored and transmitted in computer systems. Characters are stored and transmitted using the ASCII code system, which involves storing each typeable character as a specific decimal number.  You can search for "ASCII table" to see how the system works. Punctuation marks are represented by the numbers between 32 and 47,  the numbers 58 to 64,  the numbers 91 to 96, and the numbers 123 to 126. The numbers 0 through 9 are represented by the numbers 48 to 57. Uppercase letters are represented by the numbers 65 to 90. Lowercase letters are represented by the numbers 97 to 122. 

Under the ASCII system a keyboard character is represented by a decimal number between 1 and 128. Storing or transmitting a decimal number that can vary between 1 and 128 always requires at least 7 binary bits. For example, the number 100 is stored as 1100011 in binary, and the number 115 is stored as 1110001. Since it requires at least 7 bits per character to store or transmit a character, and since an English word of average length and the space character after it consists of six characters, the correct calculation for the output bit transmission rate of typing 120 words per minute is below:

Bits per minute = 7 * 6 * 120 = 5040

Bits per second = 5040 /60 = 84

So typing at 120 words per minute involves a bit output of about 84 bits per second, which is 800% higher than Zheng and Meister have estimated. The fastest human typist typed at 305 words per minute, which is a rate of more than 200 bits per second. This is an output more than twenty times greater than the "10 bits per second" figure cited by Meister and Zheng. 

But the numbers above are actually a 50% underestimation of the throughput occurring. The typists who type 120 words per minute are reading as they type. To properly calculate the throughput (the total number of bits moving around), we must calculate both the activity of reading the 120 words and the activity of typing the 120 words.  Doing that, we are left with a total throughput of 168 bits per second, which is more than 1600% higher than the throughput Zheng and Meister have estimated.

An ordinary old guy like me can sing the four lines below in about 8 seconds:

I am the very model of a modern Major-General,

I've information vegetable, animal, and mineral,

I know the kings of England, and I quote the fights historical

From Marathon to Waterloo, in order categorical;

These lines are the opening lines of Gilbert and Sullivan's patter song with the same title as the first line. The four lines have a total of 207 characters, including the spaces. I can sing them at a rate of about 26 characters per second. That is an output of about 208 bits per second. This is an output more than twenty times greater than the "10 bits per second" figure cited by Meister and Zheng.  And I haven't even considered the additional information content involved, because when I sing that each syllable is a note on the musical scale; and I didn't factor in that additional part of the information. To calculate the total throughput, you should double that figure of 208 bits per second, because what is going on is not just verbal output but also recall. When I sing that fast I am both recalling at more than 208 bits per second and also singing at more than 208 bits per second. So the total throughput should be calculated as  more than 416 bits per seconds. This is an output more than forty times greater than the "10 bits per second" figure cited by Meister and Zheng. I can also add additional throughput by dealing out playing cards in a line as I sing at the rate above. 

John Moschitta Jr. demonstrated the ability to speak at a rate of about 50 characters per second, an output rate of more than 400 bits per second. Any guitarist or pianist simultaneously singing and playing at the fastest tempo will be producing throughput of 200 bits per second or more, often more than 400 bits per second. To calculate the total throughput, you would need to start by computing three different things: (1) the bits required for each word; (2) the bits required for each played note; (3) the bits required for each note sung. After making that calculation, you should double the result, to account not just for the auditory output, but the additional work of either recalling all the information or reading it from a musical score. 

Meister and Zheng's claim about humans living at 10 bits per second is obviously nonsense that is wrong by a factor of very many times, wrong by at least 4000%, and probably very much more. Why would people want to claim that human mental performance is so many times  slower than it is? To answer that question, we may consider the general tendency of biologists to make grotesque misrepresentations of  human mental performance. Biologists and materialists are constantly trying to depict humans as mental performers vastly worse than humans are. 

In a previous post I stated a rule:

The first rule of neural explanation: the credibility of any claim that human minds are produced by brains (or are the same thing as brain states) is inversely proportional to the diversity and depth of human mental experiences, the number of mental powers that humans have, and the speed, skill and depth of such powers. 

Because of this rule, biologists and materialists adhering to the "brains make minds" dogma very frequently depict human minds and human mental performance in misleading ways, trying to depict humans minds as very much less than they are, and human mental performance very much worse than it is.  There are many examples and facets of this misrepresentation:

(1) A biologist may make the absurd claim that there is no fundamental difference between the minds of humans and the minds of other mammals, a glaring misstatement made by Charles Darwin (link) and others. 

(2) A biologist or some other materialist may depict human memory as error-prone, thereby ignoring or failing to study numerous cases of flawless memory performance by humans, such as cases of people who memorized without error poems with thousands of lines. 

(3) A biologist or some other materialist may claim the non-existence of human psychic powers, failing to study hundreds of years of very good evidence for the reality of such powers. 

(4) A biologist may depict human learning as something requiring multiple sensory exposures, ignoring abundant evidence everywhere that humans very often learn things after a single sensory experience. 

(5) A biologist or some other materialist may use the silly trick of describing the human mind as mere "consciousness," a trick that makes human minds sound a million times simpler and less impressive than they are. 

The "10 bits per second" nonsense of Meister and Zheng is just the latest example of biologists senselessly depicting the human mind and human mental performance as very much less than it is. Their very obviously erroneous paper has been trumpeted by science journalists, who acted in their usual uncritical way, rather as if they were North Korean journalists who repeat without question every claim made by North Korean government officials, no matter how obviously false it may be. 

So, for example, we have an article on the paper by Carl Zimmer, who has very long acted as a "pushover" science journalist for the New York Times, seemingly accepting without critical scrutiny very many cases of dubious or untenable claims made by professors. Zimmer promotes the nonsensical claims of Meister and Zheng, just as he has promoted very many professor claims that fail to hold up to critical scrutiny. He quotes Meister stating the glaring falsehood that "we are incredibly slow." It does not occur to Zimmer to point out that this claim is obvious nonsense, given that humans can instantly provide detailed answers and biographical descriptions as soon as they hear a single name or as soon as they see a single picture of someone's face. That is blazing fast speed beyond any possible neural explanation, particularly given all of the many speed bumps and slowing factors in brains. We recognize and recall at  the speed of souls, not brains. 

In general, neuroscientists are very often poor scholars of human minds and the best examples of human mental performance, poor scholars of human mental experiences in all their strange variety, poor scholars of brain-related medical case histories, and also surprisingly inadequate general scholars of brains themselves. A broad and very deep and thorough study of best human mental performances, human minds, human mental experiences in all their variety, human medical case histories in all their variety, and human brains (including all of their physical shortfalls) is not conducive to holding the type of "brains make minds" and  "brains store memories" dogmas that neuroscientists cling to so zealously. A very interesting topic for a scientific paper would be to do a face-to-face survey of 100 neuroscientists attempting to reveal how well they understood basic facts about human mental performance and facts relevant to whether brains make minds and whether brains store memories.  I think the poor results would be shocking.