When did the concept of temperature first arise?

Question

I am interested in when the concept of temperature first arose. This seems surprisingly hard to pin down from the web-based research I've done so far. I am mostly interested in how people thought about the concept (or its precursors) prior to the invention of thermodynamics and thermometers.

On the one hand, the idea that one thing is hotter or colder than another thing seems like such an obvious concept that one might expect it to have existed since prehistoric times. At the very least, people must have known since the invention of ceramics that a bright red glow is hotter than a dull red one, and that this will systematically affect the properties of the resulting material.

But on the other hand, it seems that thermoscopes (including those built by Galileo) were only able to measure the relative hotness or coldness of the thermoscope itself, since they weren't equipped with a scale that would allow comparisons with other objects. Adding a scale to a thermoscope doesn't seem like a technologically difficult challenge, which suggests that the concept of temperature as a numerical value possessed by all objects did not exist until the 1600s or so.

After that things seem to become a bit clearer, with the development of thermodynamics leading to a better understanding of what heat and temperature are, how they behave and how to measure them. My interest is in how people thought about the concepts before that.

Given this, I guess my questions are:

• Prior to the 1600s, how did people think about hotness and coldness?

• When did the notion arise of temperature (or hotness/coldness) as a one-dimensional quantity? (That is, the idea that for any two objects, either one is hotter than the other or they are the same temperature.) Did that only appear during the development of thermodynamics or was it known long prior to that?

• Similarly, when did the notion arise of temperature as distinct from heat? Did that only appear with the development of thermodynamics, or was there a less formal understanding of the distinction before that?

• When Galileo and his contemporaries were building thermoscopes, did they have the concept of temperature, or were they only intending to measure heat? (That is, is the reason that they didn't add scales because they couldn't, or because they didn't know it was a good idea?)

Answer 1

Your link quickly leads to Middleton, History of the Thermometer and its Use in Meteorology (1966). Pre-1600 (p. 3):

The opposition of “hot” and “cold,” like that of “dry” and “moist,” [was] used by Aristotle in the formation of his doctrine of opposites [with] no attempt to assign numbers to these qualities. The great physician Galen [c.129-216] seems to have introduced the idea of “degrees of heat and cold,” four in number each way from a neutral point in the middle. (...)

Strange as it may seem, the idea of a scale of temperature was familiar to physicians before they had any instrument to measure it with. This is illustrated by the De logistica medica [1578] of Johann Hasler of Berne. Hasler’s very first “Problem” is entitled “To find the natural degree of temperature [“temperie”] of each man, as determined by [latitude] and other influences.” (...) Hasler showed this supposed relationship by an elaborate table, in which the 9 degrees of heat in the first column and the Galenic degrees of heat and cold (...) are set opposite the latitude.

(As a review hints, J.-M. Dureau-Lapeyssonnie in Médecine humaine et véterinaire à la fin du Moyen Âge (1966, pp. 209-211) traces scales further to physicians Ricart (c.1360-1422) and Al-Kindi (c.801-873). However, this still involved four “elementary qualities” (hot, cold, dry, moist) evolving independently (heat can increase without cold decreasing), so wasn’t one-dimensional. Physicians like da Monte (1498-1551) did write temperatura, but apparently in the earlier sense “state of being tempered or mixed, later synonymous with temperament”.)

Post-1600, Middleton essentially argues that modern temperature is born the day it gets measured by attaching a scale to a thermoscope, i.e. at the same time as the thermometer (pp. 4-5):

a distinction must be made between the terms thermoscope and thermometer, in which a thermometer is simply a thermoscope provided with a scale. (...) I propose to regard it as axiomatic that a “meter” must have a scale or something equivalent. (...) The serious candidates for the honor of having “invented the thermometer” are usually considered to be four in number: Galileo, Santorio (or Sanctorius), Drebbel, and Fludd.

So his earliest quotes involving the new concept are (p. 9) by Santorio (1612, reprint 1632):

(LXXXV.X): I wish to tell you about a marvellous way in which I am accustomed to measure, with a certain glass instrument, the cold and hot temperature [“temperatura”] of the air of all regions and places, and of all parts of the body; and so exactly, that we can measure with the compass the degrees and ultimate limits of heat and cold at any time of day.

(LXXXVI.III): the temperature [“temperatura”] of the air can be observed not only in so far as it belongs to the body, but also as a thing in itself; so that the range between very hot and cold temperatures [“temperatura”] of the air can be exactly perceived. For we have an instrument with which not only the heat and cold of the air is measured, but all the degrees of heat and cold of all the parts of the body, as we show to our students at Padua, teaching them its uses; and they have heard about this novelty with no little astonishment.

and (p. 7) by Sagredo in letters to Galileo (1613, 1615):

The instrument for measuring heat, invented by your excellent self, has been reduced by me to various very elegant and convenient forms, so that the difference in temperature [“temperie”] between one room and another is seen to be as much as 100 degrees. With these I have found various marvellous things, as, for example, that in winter the air may be colder than ice (...)

I have been making additions and changes every day to the instrument for measuring temperatures [“temperamenti”] (...)

Galileo himself also used the word, e.g. in Pensieri varj (c.1619, published 1744, reprint 1855):

granted that air contained in the instrument is at the same temperature [“temperie”] as other air in the ambient room (...)

Answer 2

As Francois Ziegler pointed out, Galen introduced the idea of four degrees of heat and four degrees of cold on either side of a standard neutral temperature around 150 A.D.

By the 1300s, the Oxford Calculators, associated with Merton College, Oxford, talked about temperature as if it were a continuous one-dimensional quantity, akin to quantities such as position and velocity. The important points are described in Lindberg's Beginnings of Western Science:

The fundamental idea was that qualities or forms can exist in various degrees or intensities: there is not just a single degree of warmth or cold, but a range of intensities or degrees running from very cold to very hot...

Reflection about qualities, their intensity, and their intensification thus led the Mertonians to a new distinction: between the intensity of a quality (defined above) and its quantity (how much of it there is). An example will help us to understand this distinction: it is obvious enough in the case of heat that one hot object can be hotter than another; this is a reference to the intensity of the quality, what we call "temperature." But we also have a conception of the quantity of heat--how much of it there is. If we have two objects at the same temperature, one of them twice as big, that larger object clearly has twice the "quantity" of this quality of heat.

Nicole Oresme then went so far as to represent temperature (as well as things such as pain and grace) graphically in his Treatise on the Configuration of Qualities and Motions. Unfortunately, I can't get my hands on an English translation, but Lindberg describes it like so:

Take a rod AE... heated differentially, so that the heat increases uniformly from one end to the other. At point A and at whatever intervals you choose, erect a vertical line representing the intensity of heat at that point. If (as we have postulated) the temperature increases uniformly from A to E, then the figure will reveal a uniform lengthening of the vertical lines.

Granted, what these guys seem to have been doing was to fill up a rod with a bunch of heat almost like filling a bucket with a liquid, so that there's more heat on one end than the other, then looking at the density of heat (or "intensity of heat") at a given point, which is not really the modern definition of temperature. But then, of course they wouldn't use the modern definition of temperature, would they? Regardless, I'd say this more or less reflects their thinking about warmth and coldness, or temperature.

As for why someone wouldn't add scales to a thermoscope, I think it's a little subtle, and you have to put yourself into the shoes of someone from the 1600s or 1700s.

The problem with measuring temperature is a little different than the problem of measuring speed. Everyone knows that speed is a measure of how much distance something travels in a given time. But there are philosophical issues with measuring temperature. What exactly are you measuring? As you said, everyone has an intuitive sense of one body being warmer or colder than another. But what exactly does that mean? The modern definition,

$$ T = \partial{U}/\partial{S} $$ is obviously of no help to someone pre-1600s.

The Oxford Handbook of the History of Physics gives a good description of some of the problems. First, you have to agree on certain "fixed points". It's obvious today that something like the freezing and boiling point of water works. But is it really that obvious? Before reliable thermometers, or even a definition of temperature, how do you know water always freezes at the same "temperature"? People tried a number of different fixed points before finding agreement. (Blood temperature, temperature of deep caves, melting point of butter, etc.)

Second, once you've come up with your fixed points, it's easy enough to just divide your scale up into 100 equal-spaced degrees, but everyone's scale was different in non-linear ways! There's a reason we use mercury in our thermometers. Once you make a thermometer that is based on a material expanding, like mercury or alcohol, you've essentially defined temperature to be tied to that material's rate of expansion, but materials don't all expand at a uniform rate! In other words, you can make two "thermometers" out of different materials that agree on your fixed points, but disagree in between! How do you know which one is the "correct" one if you don't already have a reliable thermometer (or a quantitative definition of temperature)?

These details were all worked out mostly in the 1700s. I think it's safe to say that before then, people's concept of temperature was similar to Oresme's, in that it was clear that some things are hotter than others, and one thing might feel the same temperature as another, but this was all a qualitative thing, similar to the sensation of pain.

(Edited to add more details and address the valid point made in the comments that Oresme might have been talking more about heat than temperature.)