Thread by @EveKeneinan

We are going to be examining this definition of “experiment” given by @Kopernikus1966.

“Experiment: the measurement of a predicted quantity under given restraints.”

@Kopernikus1966 Now, this normally shouldn’t need to be said, but it does in this case. The purpose of a definition is to make clear what the term being defined MEANS.

This purpose is defeated if a term is defined in terms which are themselves obscure or vague.

@Kopernikus1966 I bring this up because Kopernikus is going to rely on a vague understanding of “measurement” “predicted” and even “quantity” to defend his definition.

He would really need to define THESE terms, since he uses them NONSTANDARDLY to ordinary usage.

@Kopernikus1966 The first important thing to note is that quantities are not the only things that can be sort in term of “more” and “less.”

Qualities can also. For example, one can be more or less skilled, a movie may be more or less good, a joke more or less funny.

Aristotle makes the point:

@Kopernikus1966 This is going to be important because “sorting qualities by more or less” is not “measuring a quantity” but Kopernikus is going to pretend it is.

@Kopernikus1966 Next, we need to talk about MEASUREMENT.

This is because “measurement” is actually extremely highly resistant to definition.

@Kopernikus1966 In 1932, the British Association for the Advancement of Science established a commission to define “measurement.” They worked on it for 8 years, and in 1940 published their Final Report, more or less announcing they had failed.

@Kopernikus1966 Much of the controversy surrounded the Sone Scale proposed by Stanley Smith Stevens, which purports to measure how loud a given sound appears to be to the listener.

@Kopernikus1966 Can something be “measured” which cannot be demonstrated to have a quantitative structure?

The committee thought not:

@Kopernikus1966 “…any law purporting to express a quantitative relation between sensation intensity and stimulus intensity is not merely false but is in fact meaningless unless and until a meaning can be given to the concept of addition as applied to sensation.”

@Kopernikus1966 If something does not display a quantitative structure, it cannot be measured—and one demonstrates a quantitative structure by showing that something can be ADDED.

@Kopernikus1966 On the other side, it is trivially obvious that we can, with our hearing, distinguish more and less loud sounds.

But are we then MEASURING LOUDNESS?

@Kopernikus1966 Stevens, author of the Sone scale, says yes. HIS position is this: "we may say that measurement, in the broadest sense, is defined as the assignment of numerals to objects and events according to rules.”

Note the “assignment” of numerals “according to rules.”

@Kopernikus1966 Stevens proposed a typology of measurement scales which became and still is COMMONLY ACCEPTED in the social sciences.

1 Nominal scales
2 Ordinal scales
3 Interval scales
4 Ration scales

@Kopernikus1966 A nominal scale is the assignment of an arbitrary quantity to members of a group by means of a rule.

Since this JUST IS measurement, according to Stevens, assigning numbers to football players is measuring them.

@Kopernikus1966 Obviously, one of the primary objections to nominal scales of measurement BEING scales of measurement at all, is that there seems to be NOTHING in the objects that is being measured.

@Kopernikus1966 Next is the Ordinal Scale, which means “Ordering” scale, in which we sort things out according to an order of more or less, but—listen up Kopernikus—not non-arbitrary quantities enter into it.

@Kopernikus1966 Likert Scales are typical Ordinal Scales. That is the, the things that ask you

Are you

Very Satisfied
Satisfied
Neutral
Unsatisfied
Very unsatisfied

@Kopernikus1966 I could of course write that as

5 Very Satisfied
4 Satisfied
3 Neutral
2 Unsatisfied
1 Very unsatisfied

or as

1,000,000 Very Satisfied
3000 Satisfied
19.2 Neutral
0 Unsatisfied
-1111 Very unsatisfied

@Kopernikus1966 These are what is going on when we rate a movie from 1 to 5 stars or 1 to 10 or 1 to 100.

Now, presumably these are NOT MEANINGLESS, but are we really MEASURING something here?

@Kopernikus1966 We are arbitrarily assigning numbers to qualities, in this case “better” and “worse.”

The judgment may be valid, but the quantitative aspect is unreal.

It would be NO DIFFERENT if we used NO QUANTITIES and instead used

A
B
C
D
F

@Kopernikus1966 Ordinal Scales at their best DO produce a genuine ordering, but they cannot be added, since there is no true quantity underlying.

@Kopernikus1966 So it is questionable whether ORDINAL SCALES are genuine cases of MEASUREMENT of any QUANTITY.

@Kopernikus1966 Another excellent example of an Ordinal Scale is the Mohs Hardness Scale.

A substance is deemed “harder” than another substance if the one can scratch the other and not vice versa. Some substances were picked as “benchmarks” and given numbers:

Talc is 1 and diamond is a 10.

Because that’s just the numbers.

You can’t add them.

There’s no meaningful sense in saying a level 6 hardness thing is “twice as a hard” as level 3 one. That’s not true.

Presumably hardness is a real quality of things, though.

Maybe it maps onto a quantity somewhere deeper, or maybe it doesn’t.

Next we have Interval Scales. Here, ORDER between the variables is known (like the ordinal scale) AND the DISTANCE between any two is constant.

In color, HUE is expressed on an Interval Scale.

The best known Interval Scales are the Fahrenheit and Celsius temperature scales.

In Fahrenheit, freezing to boiling is 32˚ to 212˚. In Celsius 0˚ to 100˚. They measure the same thing, but the INTERVAL is different.

The reason Fahrenheit and Celsius are Interval Scales is that they have an ARBITRARY ZERO.

To have the best kind of scale, a Ratio Scale, one needs a ACTUAL ZERO.

If there were such a thing as ABSOLUTE ZERO, then we could have a RATIO SCALE of heat.

Oh, there is and we can!

Which bring us to Kelvin.

It was unknown if temperature could EVER be measured as a ration scale or whether it was confined to an interval scale (as HUE seems to be).

Turns out there’s an absolute zero.

Same with length, etc.

Most measurement in the physical sciences and engineering is done on ratio scales: mass, length, duration, plane angle, energy, electric charge.

Ratios are VERY USEFUL because having a non-arbitrary zero point makes it meaningful to say e.g. that one object has "twice the length" of another.

Ratio Scales can be described as specifying "how much" (i.e. an amount or magnitude) or "how many" (a count).

Interval scales can only do this RELATIVE to something else.

Ordinal scales can only say “more” or “less” RELATIVE to something else.

Nominal scales say nothing.

So, the question arises: what is it to MEASURE something?

Most people, when they think of measuring, think of determining a definite quantity of how much or how many as per a ratio scale.

In the Olympics we can use an Ordinal Scale to “measure” athletic performance.

1st place, 2nd place, 3rd place, etc.

Now, this is a serious problem of DEFINITIONS.

In psychology and the social sciences generally, Stevens’ definition of “measurement” has prevailed. This is likely for self-serving reasons.

Psychologists and social scientists generally claim to be “measuring” something so long as they are able to “assign quantities to something according to a rule” (Stevens’ definition).

But this makes them disregard whether that which these quantities are being assigned HAS A TRUE ADDITIVE STRUCTURE.

I emphasize sorting of things into qualitative more and less, better and worse, is neither unreal, nor subjective, nor arbitrary—but neither is based on the additive precision of QUANTITY.

QUALITIES cannot be measured with precision since if they could, they’d be QUANTITIES.

This is DEATHLY important to understand, because ALL the basic elements of ETHICS are QUALITIES, not QUANTITIES.

The fact that RATIO SCALES and thus QUANTITIES are central to the natural sciences has misled generations now into thinking “ethics is subjective.”

What are the QUALITIES of ethics?

Good and evil, good and bad, better and worse, right and wrong, just and unjust, fair and unfair, wise and foolish, virtuous and vicious.

NONE of these can be MEASURED WITH QUANTITATIVE PRECISION, for the simple and sufficient reason they are not quantities.

Qualities, by the way, are more important that quantities, although quantities do have that rational exactness that we like so well. Aristotle warns about it. Mathematics, he says, “flatters the intellect.”

Descartes and much of the early modern thinkers are victims of this. Most of them were mathematicians and it went to their swelled heads.

Someone like Kopernikus is EVEN NOW telling me that qualitative moral judgments are “subjective.”

I say that child rape is A WORSE CRIME, a MORE EVIL DEED, than insulting someone.

I also say that my knowledge that this is true is stronger than any argument Kopernikus makes.

I would like to SEE him make the case that e.g. petty theft, say, shoplifting a Snickers bar from a 7-11 is MORE EVIL or MORE WRONG or WORSE than the violent rape of a child.

I think it would be funny.

Morbidly so.

But it is interesting how far some moral subjectivist will deny reality.

HOW MUCH WORSE “violent child rape” is than “shoplifting a Snickers bar” no one can say with EXACT PRECISION.

Because “worse” is a QUALITY.

The most we can say is “A hell of a lot worse.”

But this is OBVIOUSLY not subjective.

A judge who sentenced a man who violently raped a child to the punishment normally reserved for shoplifting a Snickers bar wouldn’t just be WRONG but doing judicial malpractice, and should be REMOVED from the bunch.

And we know this.

The only people who can think moral subjectivism is remotely rational are privileged children (e.g. many freshmen) and twits on Twitter who have never experienced an act of evil.

Unfortunately, DOING EVIL to such a person, while it would cure them of their subjectivism, is unethical as a form of pedagogy.

I can relate an incident of a former colleague of mine. She was, as all teachers of freshmen, running up against the “it’s all subjective, man” types in her class.

One of her students was a Rwandan transfer student. He had lived through the Rwandan genocide. He was a Tutsi. After a few of the students had expressed how sure they were that “ethics is subjective,” he raised his hand and my colleague called on him. He didn’t speak for a time.

He then related how he was alive only because he had hidden inside a large stack of corpses when the Hutu came back to his village the second time, and hacked his family apart with machetes as he watched from his hiding place, unable to move or make a sound.

He stared at the floor the whole time he told his story. He then looked up and said “There is EVIL.”

And looked around the room.

No one could meet his eyes.

That ended all “ethics is subjective” nonsense in that particular class.

Let’s get this back on track though.

We’ve gone for quite a ways on this thread.

In 1747, James Lind was working on the treatment of sailors with scurvy. He took six groups of scorbutic sailors. They were all given the same diet, except each group was given an extra: cider, sulfuric acid, seawater, vinegar, citrus fruits, and barley water.

His idea was that acid might cure scurvy.

He found that CITRUS FRUIT did.

This is considered the first clinical trial in history.

Was it an experiment? I would say it obviously was. Not only was it in experiment, it made a hugely important discovery in the prevention of scurvy, which “killed more British sailors than the French and Spanish combined.”

Was it “the measurement of a predicted quantity under given restraints.”

Nope. No quantity was involved, not was anything predicted. It was a “try this and see what happens.”

It was a trial under controlled circumstances. That much is true. Experiments do try to ISOLATE irrelevant factors.

But the finding was just the empirical discovery “citrus fruits cure scurvy.”

That scurvy is caused by lack of vitamin C and citrus fruits are full of vitamin C wouldn’t be known until much later.

“Citrus fruit is efficacious in the treatment and prevention of scurvy” seems like an important empirical discovery made as the result of an experiment: Giving scorbutic sailors different things to eat.

No quantity. No prediction.

How about Pavlov’s dogs?

Pavlov hypothesized that he could get dogs to associate a bell with dinner. You all know the story: he ran a bell when the dogs were fed, and after a while, when he rang the bell, they’d salivate in anticipation of being fed.

This is probably one of the most famous experimental results of all time. So much so that you all KNOW ABOUT IT already (bet you didn’t know the scurvy thing!).

Again. No quantity and no prediction.

Actually, Pavlov hypothesized he could get the dogs to link the bell to food, so I guess that is a prediction, but hardly of any quantity.

A behavior isn’t a quantity.

The Little Albert experiment is fun too. Talk about unethical psychology. Little Albert was a baby that a psychologist made afraid of a white rat. The baby, Albert, didn’t fear the rat at first, but the psychologist kept making a loud noise, firing a gun or something.

As predicted, the baby associated the white rat with the scary loud noise and became afraid of the rat.

We could go on endlessly with these, from Albert Bandura and his Bobo dolls to Milgram and his Obedience experiments. Psychologists do TONS of experiments where they sort out connections and relations of more and less, but very little MEASUREMENT.

How can you measure the psyche?

If you insist that experiments MUST involve MEASURE of QUANTITY, you wipe out pretty much all experiments is the social sciences, which are NOT USELESS, and more importantly, NOT NOT EXPERIMENTS.

Or you have to contort reality and say that “more” or “less” is REALLY quantity even though you can’t determine any value of it.

Milgram showed some people are more susceptible to authority than he expected. A lot more.

But there’s no SI unit of “obedience.”

I just felt like talking about things, like Ordinal Scales and Scurvy and what I consider to be one of THE most important things ANYONE can know:

1 Qualities are not Quantities.
2 That doesn’t make them subjective.
3 We can judge better and worse.
4 This cannot be PRECISE.

So I’m going to wrap of this long ramble now.

But bear in mind that ONE POINT if nothing else.

This is FALSE, FALSE, FALSE:

"If it isn’t a measurable objective quantity then it must be subjective."

The Holocaust was more evil than a case of shoplifting a Snickers bar. You can know that with as much CONFIDENCE as you can know anything.

If a person says you have to produce a quantitative proof of that before he believes it, call the police.

As philosopher David Stove used to say,

Footnote: I found a nice page explaining or at least outlining the four types of scales.

These are important to know, so here: questionpro.com/blog/nominal-o…

Remember: there is ongoing dispute whether all four are “measurement”, only 2-4, only 3-4, or only 4.

I lean to “only 4.”
That is “no ratio scale, no measurement.”

But rank-ordering is a real thing, even if it isn’t measuring.

And here’s a paper I thought valuable:
psychology.okstate.edu/faculty/jgrice…