Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
In Defense of Fahrenheit (lethalletham.com)
28 points by johnmyleswhite on May 27, 2019 | hide | past | favorite | 90 comments


The author argues against Celsius because it 'feels' wrong? Well, I've grown up with Celsius all my life, and Farenheit just 'feels' wrong to me.

Everything is relative. The author deems 35 deg C as being "grossly hot", but that is just the average temperature to me where I live. He calls 0 degrees not even cold, but for me, that is almost unbearably cold.

Having a datum of 0 for freezing makes sense, especially for professions like flying (I used to be a pilot). Knowing that you are close to a point when water will turn from liquid to solid can be life saving, and I have no problems with using negatives to determine HOW solid.


The "intuitive" argument comes up a lot in relation to units of measure, and it just makes me wonder how people can have such lack of self awareness.

Anything that is familiar is easier, that tells you nothing about the thing itself. This seems so obvious to me I wouldn't even think of basing an argument around it. Is the concept of subjective experience that hard to imagine for some people?


I mean no disrespect in saying this, but i genuinely think Americans don't like to leave their comfort zone too often and it results in a lot of subjective bias.

I read something like 40% of Americans have never left their country, and about 10% have never left their own state. That means a great deal of them have never experienced a temperature outside of their own.


No disrespect taken!

I think you may be intuitively underestimating the size of the US. I live in Arkansas, which is a fairly mid-sized state. Driving across it at highway speed (70mph/115kph) E/W or N/S takes about four hours. It’s slightly larger than England in area.

Consider someone who lives in Austin, TX. They could drive four hours in any direction and still be in Texas. The US as a whole is approximately the size of all of Europe.

Given that kind of scale, it seems illogical to say that many Americans have “never experienced a temperature outside of their own” merely because we tend to not travel internationally as often as others.

As for comfort zones - I think that’s something of a funny thing. I’ve been to Montreal a couple of times, and honestly felt more at ease there than when visiting New York City, Los Angeles, or Hawaii. There is a huge amount of diversity in terms of culture within the United States.

The other big factor is expense. My wife and I plan to visit Scotland and Italy at some point. We estimated that a week in Scotland for the two of us would cost about $7,000 - $5,000 of which is airfare alone! Consider that the median household income in the US is about $60,000 ($50,000 after federal income taxes). That week in Scotland is approximately the equivalent of two months’ income for a median household! In comparison, a round-trip flight from Heathrow to Istanbul is $330. It’s much, much cheaper for Europeans to travel internationally.


On the similar line: Americans are often said to be unknowing since they mix up European cities / countries and can’t point them out on a map. But the reverse is totally true as well. Ask Europeans where Arkansas is (or how it is pronounced;) and few will know the answer. This mostly boils down to what you learn in school and how much ‘local’ news you are exposed to. I had plenty of European geography in school and very very little US geography. Of course I’m not talking about US/European millennials as they seem to know nothing about geography at all ;-)


> never experienced a temperature outside of their own

What does that even mean? You can experience temperatures ranging from 0°F to 100°F within six months in New York City. Why would you need to travel to experience different temperatures?


> The "intuitive" argument comes up a lot in relation to units of measure, and it just makes me wonder how people can have such lack of self awareness.

As someone who has grown up with a mix of various Imperial (not US) measures and metric ones, I often find some more intuitive than others even when I am familiar with both.

The most obvious his human height. A tall man is about 6' or more. Most people are some number of inches less than that. Moreover, the inch and not the centimetre is roughly the amount at which you would casually notice two people were slightly different in height.

On the other hand, for horizontal distance, metres (or yards) make more sense than feet. To the extent that when American tell me how far away something is in feet, I just multiply by three before trying to use the info.

I think weight in stones is similarly better than either kilos or or pounds, but I am personally more familiar with kilos.


> On the other hand, for horizontal distance, metres (or yards) make more sense than feet. To the extent that when American tell me how far away something is in feet, I just multiply by three before trying to use the info.

Wouldn’t you divide (rather than multiply) by three to get the distance in yards, which would be closer to meters?


No, when people say "The thing is N feet down the road", I find that it is really N yards down the road. That is, I multiply the distance by three.

Though it's true, the alrgorithm for doing the multiplication happens to is `s/feet/yards/`


My argument is not that all measurement systems are equally ergonomic, it's that you can't reach any conclusions about ergonomics by examining your subjective experience of a system you grew up with and use every day vs. one you don't.

This is what I though would be blindingly obvious, but apparently isn't.


This, the author is a phd surely he should see his own bias here?

"I found it annoying to use negative numbers for temperatures that aren't really that cold, and then have 35℃ be grossly hot." living in Sweden i've always used 0 as the "feeling" middle point, temperatures usually ranging between -20 / +30 C, feels very natural.

That saying the task seems quite fun, would be nice to replicate it coming from a celcius background.


0 is also very important for driving, camping etc. Water freezing is just so important point.


Memorizing 32 is not much different than memorizing 0, despite how obvious it seems to the centigrade folks.


I have lived under both regimes, and it's easy to adjust to either one... It's just a matter of a few years.


Do you have a preference for one vs the other?


Nope. Not at all. You have references in your mind for what a certain number correlates to and that's all that changes...


The temperature scale is a) rather arbitrary b) rarely converted in everyday life.

So it's the least bad american unit. If I was suddenly forced to switch from C to F I'd just adjust. It's not that bad.

The others such as distances and volumes are insane and of course not because of the units but because of being non-decimal. Example: figuring out the volume of a swimming pool given its dimensions. Which length unit would you measure the pool in, in order to have a calculation for the volume (in gallons) that doesn't require a calculator? (A: a unit that is 6.14 inches long)


Quite. He's solving a nonexistent problem. Next he'll tell us that English is H. Sapiens' natural language.

> I found it annoying to use negative numbers for temperatures that aren't really that cold

All Australians understand perfectly well that anything below 20C (or 25 above the Tropic of Capricorn) is cold.


It's not so much 0 that's the problem with Centigrade for weather purposes, it's the 100. 100 F is rather warm. 100 C is DEAD.

I don't have a problem with Centigrade for other stuff, but I think Fahrenheit captures the human ambient environment range better.

Disclaimer: American, where I remember the vain national attempt at metrication as a young child and still have to convert everything mentally when visiting my Australian in-laws


But I still find the 100 a useful demarcation - especially when cooking. I find it handy to know how far away my soup or tea water is away from boiling without doing complicated maths in my head.

The equivalence to the percent scale also makes it easier to conceive - i.e. if I see my kettle on the hotplate registering 75 degrees C, I immediately think "Oh, it is three quarters of the way to boiling".


You have provided a fine example of why Celsius is bad when you think "Oh, it is three quarters of the way to boiling". No, 75 degrees C is not.

Your calculation: 75/100 giving 0.75 which is "three quarters"

Proper calculation: (75+273.15)/(100+273.15) giving 0.933 which is "fourteen fifteenths"

Your tap water is probably more than three quarters the way to boiling.


You could ask: "Three quarters" of the way from WHAT temperature to boiling?

It sounds like the grandparent is trying to measure how far through the task of boiling their water is, by interpreting degrees Celsius as percentage done.

Maybe they usually make soup or tea by boiling ice cubes. (Perhaps they're Siberian?)

Are they saying that 75 degrees mean three quarters of the way from room temperature tap water to boiling water? Or are they talking about boiling an ice cube, so it's three quarters from 0 Celsius freezing?

Interpreting degrees Celsius as a percentages is not an intuitive way of measuring time elapsed to boil tap water, because 75% between freezing and boiling doesn't mean 75% of the time required to boil has elapsed, since you're starting from room temperature when you boil tap water, not freezing.


> Fahrenheit captures the human ambient environment range better

It doesn't. It's just a number which denotes a temperature. You'd quickly get used to celsius if you switched to it.


The point I'm trying to make is that if 0-100 is a useful mental range, and people seem to be arguing that it is given boiling and freezing points, then for the range of temperatures humans can reasonably occupy the scale should be as close to that range as possible.

This is different from an intuition argument. My wife, an Australian, finds nothing intuitive about 72 F as room temperature, just as I don't find anything intuitive about 22 C. You're right I'd adapt if I had to, but no one lives anywhere for very long at 72 C.


> for the range of temperatures humans can reasonably occupy the scale should be as close to that range as possible.

Why? I don't think I'll die at 110 degrees Fahrenheit and that's off the scale. And I routinely make use of things which get hotter than I can live in. I don't live in my kettle or my oven.


Sure, but you will die at 80 C, and that isn't off the scale. So which mapping is closer?

Again, I'm making my argument about weather. Clearly for other arbitrary temperatures this doesn't apply. If you're actually arguing for a single temperature scale for cooking, weather and anything else, then I think most people will end up concluding it'll be what you're most used to working with.


The point is there's no need for any mapping to "livable conditions". Just use Celsius and be happy. The rest of the world does. It works.


Sure, but you will die at 80 C

Eventually. 80 C is normal sauna temperature.


But surely Americans use temperature for cooking too and not just for weather, and then 100°C is boiling which is also at least somewhat useful to know.


When cooking I have never needed to know what temperature boiling water is. The fact that it's boiling is enough. Never mind that water only boils at 100°C at sea level and most Americans don't live at sea level.


I guess prioritizing cooking or weather is what it, er, boils down to. ;)

If folks want a single temperature scale that works in a fry pan, around Moscow and on Proxima Centauri, there's always Kelvin ...


The argument that 0-100 being an intuitive and useful daily use scale is actually a good one.

However, everything else falls flat.

Nobody actually uses decimals with celcius, because nobody can tell the difference between 8.2 degrees and 8.7 degrees. It's just 8 and 9.

Zero being freezing is a critical number to know, because that's when you can get black ice (+/- wind-chill) and snow.

Knowing that 35 is grossly hot is no more difficult than knowing -32 is freezing.

Celcius has simple ranges: Below 0: it's freezing, bundle up 0-10: it's cold outside, wear a jacket 10-20: good for physical activity, sweater and light jacket required 20-30: it's comfortable outside 30+: stay in the shade

That said, all our appliances (oven, barbecue) use Fahrenheit and I have no idea what temperature to cook at in celcius.


> That said, all our appliances (oven, barbecue) use Fahrenheit and I have no idea what temperature to cook at in celcius.

It surprised me when we recruited a developer from Canada, and on his second day at work he asked what temperature he should have set the oven to. I learned that Canada has appliances labelled in "American".

180°C is the most common oven temperature required, outside 160-200°C is unusual for normal cooking.


Canadian here, 325-450F is the typical range I've needed to use, pizzas are often what I cook at those hotter temps. That's about 160-230C.

There's a surprising inconsistency in Canada for units because of our proximity to the US. For body weight and height, nearly everyone uses lbs and feet/inches, but official records (on my driver's license for example) uses kg and cm. Like you said, most appliances are F. Distances are in km but property sizes are still often in sqft (because lots of products come from the US). The list goes on.


It seems that in the extreme ranges of numbers (150+), the propensity for errors or confusion between the two scales gets riskier .

I was having an online discussion with someone in an electronics forum the other day, and we were talking about the ideal temperature for a soldering iron for a contact point using a particular type of solder and potentiometer.

He kept talking about setting the iron at 300 degrees. Now, my iron DOES have a setting up to 400 degrees, but in Celsius. He was talking in Farenheit, and had I not double checked with him, I would have fried my electronics to a crisp.


Do dial temperatures actually correlate to internal temperatures accurately at all? From what I've seen it differs wildly from oven to oven and you're better off just learning where the low, medium and high spots are.


Ive always thought ovens could do fine with a five setting control.

Warming (180-200F)/90C

Low 300F/150C

Med 350F/180C

High 400F/200C

Broil


A High of 400/200 seems quite low. I'd definitely want a Very High of ~230-250 as well.


350 to 375 is pretty significant of a gap over several hours.


outside 160-200°C is unusual for normal cooking.

Huh? I probably do more than 50% of my cooking outside that range.


In an oven?

I gave the temperatures based on my experience of cooking "normal" things, like roast meat, roast vegetables, chips, bread and cake.

They happen to correspond to Wikipedia's idea of "Moderate" temperatures:

https://en.wikipedia.org/wiki/Oven_temperatures


Sure. Roast vegetables and bread for example I almost always do in the 200-250 range. Roast meat and fish I often do between 85-150, or 220-250 if I'm finishing them in the oven. Sometimes I'll start at 250 and then finish at 150.


225°C is the standard for most frozen pizzas, fries, pretty much anything, really.


My temperatures are for a fan-assisted oven, you can add 20°C for a 'normal' oven.

Although, I've never owned an oven without a fan-assist.

(And for what it's worth, Tesco's frozen pizza and chips should be cooked at 180°C and 200°C respectively, or 200°C and 220°C without a fan oven.)


Some pastries go even higher than this, like 250° C


> Zero being freezing is a critical number to know, because that's when you can get black ice (+/- wind-chill) and snow.

For those of us who don't live somewhere that ever gets black ice, this is not a critical reason.

Also, I suspect that wind-chill does not work how you think it works.

> That said, all our appliances (oven, barbecue) use Fahrenheit and I have no idea what temperature to cook at in celcius.

Spare a thought for those of us in enlightened civilisations that have to deal with myriad north-american based cookbooks ... but having said that, it'd take you a couple of weeks, with maybe a post-it note in the kitchen for the duration, to get the hang of the conversion rate.


That's OK, cooking in the US is weird - they use cups in their recipes that are a different size from those used in the rest of the world - because of course they use fluid ounces/gallons/etc that are a different size from Imperial cups/flu ozs/gallons, much less different from a standard 250g metric cup


To provide a counterexample to your statement "nobody actually uses decimals with celsius", when I switch the climate control system in my German-made car to Celsius mode it does use decimals: either .0 or .5, and the toggle moves the temperature up or down by half a degree.

Whereas in Fahrenheit mode the toggle moves the temperature up or down by a single degree, which seems cleaner and simpler to me (not to mention the F display uses 2 digits, while the C display uses 3).


> the F display uses 2 digits, while the C display uses 3

The °F display must use three digits as well when it's more than 37°C, which isn't that rare (at least for me, maybe not where you live).


I don't think the climate control system in any car can be set to pump out 100F+ heat. Usually it ranges from 60-80F.


Oh right, I thought it was for the external temperature sensor.

Yeah, climate control won't go that far of course. It also does use half degrees (only between 19°C and 23°C in my car).


The decimal point wasn't brought up against celsius specifically.

I think the biggest factor of c/f preference is just what you grew up with and are most accustomed to using every day.


There are objective landmarks outside of culture. The absolute zero, water freezing, water boiling, human normal body temperature, etc. It's better to use a unit where important ones are aligned to small practical. Probably something better than celcius could be designed.


Completely agree.


The author lists the importance of a global temperature scale, then shows a map of North America overlaid with some dots.

> For many people that's Celsius, but for many others it's Fahrenheit.

That second 'many' is disingenuous. For about 90% of the planet's population it's Celsius, and for about 10% it's Fahrenheit.

Author also has some pretty arbitrary and bewildering requirements, including wanting a scale of 0 to 100, yet also avoiding any 3-digit numbers.

Any claims about 'intuition' are misguided -- Celsius makes more sense because a) most people use it, and b) it maps onto the rest of the SI units.

That I happen to feel comfortable at a somewhat arbitrary 21 (c) is not a sufficient or satisfactory reason for other people to let go of deprecated temperature scales.


The requirements list reads as a description of farenheit, which the author then uses in a roundabout way to conclude his presupposition that farenheit is the better unit.


> For about 90% of the planet's population it's Celsius, and for about 10% it's Fahrenheit.

I thought only the US and a few very small countries use Fahrenheit, which is not more than 5% of the world population?


A quick glance at the numbers on Wikipedia puts it a bit under 5%, the bulk of which is inhabitants of the US, but I did leave out the Bahamas for convenience.


Celsius is much more useful than stipulated here. For example, it takes 1J (Joule) to heat 1g (which is also 1ml) of pure water 1 degree Celsius.

The metric system is all weighted and related from temperature to mass to distance. This is what makes it so useful for Science, and frankly, every day use.


Please provide an example where you need to convert between Joules, grams, and Celsius in "every day use"


Estimating how long it will take the kettle to boil :-)


One calorie, not one Joule. Water has a specific heat of ~4.2 kJ/kg.


> The truth is, outside of a chemistry lab I don't care that much about the freezing and boiling points of water

What? Freezing point of water is one of the most important things with temperature outside. Will it snow or rain? Will the lake freeze or melt? Will the road be slippery or not. That is like... the most important thing about temperature and outside.

However, I agree that the boiling temperature of water is not much used outside in a daily life.


Or we could just stop this Freedom Units bullshit and be like normal people. It feels wrong is not a reason to keep doing the stupid thing.


There is a desperate attempt by some news papers in the UK to bring back Fahrenheit, The problem being only those over 40 really knows what it means to have a 85f "scorcher"

its only being done as a symbol of "freedom" from the EU loving french.


>The truth is, outside of a chemistry lab I don't care that much about the freezing and boiling points of water

Even in a chemistry lab, Celsius isn't that great, it's Kelvin that actually has an advantage. And Rankine would be about as good.


C goes 90% of the way there because 90% of temperature usage in science is for delta-T. That said, you need to be aware of it so you don't accidentally use the occasional absolute temperature in C at some point.


Yes, C is useful in science because it's commonly used, not really because of any inherent usefulness. The network effect is the usefulness, and if F had that network effect it would be just as useful.


A Kelvin temperature is just the Celsius temperature +273°.


A Rankine temperature is just the Fahrenheit temperature +459.67°.


I care very much about the freezing point of water outside. It means ice on the roads.


0F is also works for that application too in a slightly different way, it is when a salted road starts to freeze back and becomes slippery again.


My point was about in the chemistry lab, not outside.


I don't work in a chemistry lab. But why is the boiling point of water not important there?


It is important, but more important is having absolute 0, which Celsius doesn't have.

The other problem with Celsius and the boiling point is that the boiling point changes depending on air pressure, which changes depending on elevation. So it's not always 100°C.


Celsius is pretty easy to wrap my head around, even though I think in Fahrenheit. I think my main argument against Fahrenheit is when I'm cooking and look up food safety information I usually convert to Celsius in my head because I have a pretty intuitive sense of temperatures as a percentage of the boiling point of water but Fahrenheit I really don't. I also don't imagine anyone using Celsius as their "native language" routinely converts to Fahrenheit in their head just to get a better understanding of what a temperature means.


In practice a difference of 1 degree celsius is slightly noticeable. But given that 1 degree Fahrenheit is only 5/9 degree celsius, do you really care if it‘s 74 or 75F outside?


I can definitely feel the difference between when I set my car to 72F vs 73F (or any pairing of consecutive degrees F)


For everyday use the unit does not matter at all. We could drop numbers completely and just say it's freezing, cold, chilly, ok, warm and hot and be done with it. For cooking we could just use simmer, boil, stew and bake.

But Celsius interacts with other units nicely. So when you actually need some precision, Celsius works nicely. Kelvin is of course the one to choose when talking about scientific matters, but Celsius works well enough for everyday things and when precision is needed.


So why is the metric system based on water, instead of liquid hydrogen? Why not use the most common essential element, instead of some arbitrary compound?

Metric seems pretty anthropocentrist to me, since humans are 50-60% water, while the universe is 75% hydrogen.

In the Universal Hydrogen Metric system, 0 °H would be -259.2 °C, and 100 °H would be −252.87 °C.


I think there are a lot of similarities between F and C as there are with the use of military time. In the US we use AM/PM for civilian use but in England you see them use military time for things like TV show times, and train/bus times. It all seems very regimental.


I think a lot of people here are missing the most important point, which is that Fahrenheit roughly maps 0-100 to the most common temperature ranges experienced in daily weather, which is by far the most frequent use of temperature for most people.


Rankine is even better. (like Fahrenheit plus 459.67°)

You get the same step size as Fahrenheit, so you can adjust your house temperature without a decimal. (some of us care) It is also good enough for industrial food processing without that decimal, for example in canneries.

You get a correct zero. Fahrenheit and Celsius are just wrong. Think about the meaning of something like "twice as hot", which is a thing people say. With a correct zero, there is no need for negatives unless you are doing quantum physics research. If you aren't convinced, imagine we had a weight scale that used 0 for a very small adult and 100 for a very large adult.

All the normal temperatures people deal with, including oven temperatures, are exactly 3 digits. The 3-digit numbers 100 to 999 are like Fahrenheit from -359.67° to 539.33°.


I think alternative measurements are cute and I encourage their usage simply because they make epic reddit science guys so irrationally mad.

The year is Reiwa 1, it's 51° F out, and I weigh 9.92 stone. Deal with it.


LOL. Do people really still use fahrenheit over there?


More imperial entanglements than you can possibly imagine.


American customary units are not the same as imperial units. The British imperial reform in the early 1800s altered the volume measures, whereas the US continued using the Queen Anne wine gallon and the Winchester grain bushel.


tldr; Fahrenheit maps the subjective feeling of very cold to very hot to a scale from 0 to 100.

Celsius on the other hand gives you additional information about the state of water at this temperature (0C frozen, 100C boiling).

The author didnt talk about the correct physical unit, Kelvin, which tells you about the possibitiy of the temperature (no Kelvin values below 0).


Fahrenheit seems to be a good scale for weather. It would be a pretty simple matter to make a human friendly scale based on Celsius - halve the unit. Freezing is zero degrees, and boiling is 200 degrees. I call this unit Celinheit:

Here's how that looks when compared with Fahrenheit:

  0F == -35 Ch
 10F == -24 Ch
 20F == -13 Ch
 32F ==   0 Ch
 40F ==   9 Ch
 50F ==  20 Ch
 60F ==  32 Ch
 70F ==  42 Ch
 80F ==  53 Ch
 90F ==  64 Ch
100F == 76 Ch 110F == 86 Ch

The advantages of this system:

* 40-60 would be considered about ideal.

* Each 10 degree "band" has a feeling, similar to Fahrenheit.

* Converting to metric would be really easy.

Disadvantages:

* https://xkcd.com/927/


> The advantages of this system: > > * 40-60 would be considered about ideal.

Why would that be an advantage? I consider about 20–25°C ideal. Why is it worse than some other range?

> * Each 10 degree "band" has a feeling, similar to Fahrenheit.

As a user of the celsius scale, each 10 degree "band" already has a feeling to me.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: