5 ms·

> the so-called normal distribution I absolutely agree with SciAm here. Words matter. They may not break my bones, but they matter. The word "normal" is load

> the so-called normal distribution I absolutely agree with SciAm here. Words matter. They may not break my bones, but they matter. The word "normal" is loaded. The distribution was named "normal" in a different era when "scientists" believed that the distance from the tip of the one's nose to one's forehead was a measure of intelligence. It's called "normal distribution" for a reason. With hindsight and progress we can now discount the nomenclature, and clarify it. It's okay.

adhesive_wombat 5y ago
Apparently not: "Gauss himself apparently coined the term with reference to the "normal equations" involved in its applications, with normal having its technical meaning of orthogonal rather than "usual"." (Wikipedia) The interpretation of "normal" to mean "common" came later, and even then refers to the distribution itself being "the norm" rather than the distribution being about comparisons to a hypothetical "normal" object.
buntsai 5y ago
Have you actually looked up whether Gauss called this distribution "normal" because the other (parts?) of the distribution were "abnormal"? Surely it is worth at least have a quick read of wikipedia or doing a Google search before saying something that may sound preposterous...
oxfordmale 5y ago
The term normal was actually coined by Karl E. Pearson, as there there was international disagreement about the naming of the Laplace-Gaussian (or Gaussian-Laplace) distribution. "Normal" also doesn't mean "usual", and is a mathematically construct. The problem with the default person is actually that there are minor genetic differences between different ethnicities. This means that certain medical conditions are more prevalent in some ethnic groups than others (for example lactose intolerance) and this is not accounted for by assuming a normal distribution, or "a default person".
- tgv 5y ago
  I don't think it was Pearson who coined the name. Wikipedia dates it back to 1875, when Pearson was 18. Its more likely (to me), that its use grew organically, and is associated with Pearson because of his influence. > this is not accounted for by assuming a normal distribution Often, groups with different averages produce a normal distribution when combined. So it may be the right distribution, but not the proper conclusion.
tgv 5y ago
If words matter, reading and understanding should matter even more.
gspr 5y ago
> The word "normal" is loaded. The distribution was named "normal" in a different era when "scientists" believed that the distance from the tip of the one's nose to one's forehead was a measure of intelligence. No. The word "normal" is overloaded. Seriously, anyone who has ever done any amount of mathematics will have noticed this! To wit, https://en.wikipedia.org/wiki/Normal#Mathematics https://en.wikipedia.org/wiki/Normal#Mathematics . Among research mathematicians it's even an established in-joke: – Oh, so if we define X to have these properties, everything works out! – Indeed, we need to give these properties a handy name. – How about saying X is normal? – LOOOOL
timwaagh 5y ago
Agreed it's loaded. I don't like the word normal either. I think scientists should definitely not assume normal distribution. But I think the name is to put emphasis on how many things are distributed in such a manner. If you know something has a different distribution that is valuable information. You now know it can not consist of many independent similar variables being summed. So either there are not many similar parts or they depend on one another. So in that sense the name is appropriate. But I would not have any problem to rename it the common distribution or something which also conveys the point.

adhesive_wombat 5y ago

Apparently not: "Gauss himself apparently coined the term with reference to the "normal equations" involved in its applications, with normal having its technical meaning of orthogonal rather than "usual"." (Wikipedia) The interpretation of "normal" to mean "common" came later, and even then refers to the distribution itself being "the norm" rather than the distribution being about comparisons to a hypothetical "normal" object.

buntsai 5y ago

Have you actually looked up whether Gauss called this distribution "normal" because the other (parts?) of the distribution were "abnormal"? Surely it is worth at least have a quick read of wikipedia or doing a Google search before saying something that may sound preposterous...

oxfordmale 5y ago

The term normal was actually coined by Karl E. Pearson, as there there was international disagreement about the naming of the Laplace-Gaussian (or Gaussian-Laplace) distribution. "Normal" also doesn't mean "usual", and is a mathematically construct. The problem with the default person is actually that there are minor genetic differences between different ethnicities. This means that certain medical conditions are more prevalent in some ethnic groups than others (for example lactose intolerance) and this is not accounted for by assuming a normal distribution, or "a default person".

tgv 5y ago
I don't think it was Pearson who coined the name. Wikipedia dates it back to 1875, when Pearson was 18. Its more likely (to me), that its use grew organically, and is associated with Pearson because of his influence. > this is not accounted for by assuming a normal distribution Often, groups with different averages produce a normal distribution when combined. So it may be the right distribution, but not the proper conclusion.

tgv 5y ago

If words matter, reading and understanding should matter even more.

gspr 5y ago

> The word "normal" is loaded. The distribution was named "normal" in a different era when "scientists" believed that the distance from the tip of the one's nose to one's forehead was a measure of intelligence. No. The word "normal" is overloaded. Seriously, anyone who has ever done any amount of mathematics will have noticed this! To wit, https://en.wikipedia.org/wiki/Normal#Mathematics https://en.wikipedia.org/wiki/Normal#Mathematics . Among research mathematicians it's even an established in-joke: – Oh, so if we define X to have these properties, everything works out! – Indeed, we need to give these properties a handy name. – How about saying X is normal? – LOOOOL

timwaagh 5y ago

Agreed it's loaded. I don't like the word normal either. I think scientists should definitely not assume normal distribution. But I think the name is to put emphasis on how many things are distributed in such a manner. If you know something has a different distribution that is valuable information. You now know it can not consist of many independent similar variables being summed. So either there are not many similar parts or they depend on one another. So in that sense the name is appropriate. But I would not have any problem to rename it the common distribution or something which also conveys the point.