3 ms·

> "In general, air pollution makes objects appear hazier, makes it harder to identify their edges, and gives the scene a whiter tint, because pollution reflects

> "In general, air pollution makes objects appear hazier, makes it harder to identify their edges, and gives the scene a whiter tint, because pollution reflects visible light of all wavelengths," Albright said. > The team looked for these two metrics, edge strength and whiteness, in the paintings — by converting them into mathematical representations based on brightness — and then compared the results with independent estimates of historical air pollution. This and some assumptions drawn from letters by the artists that don't directly confirm anything. This study--conducted by climate scientists--shows impressionists know how to make something look hazy. The conclusion they draw sounds reasonable, but the study itself doesn't prove causation Previously not discussed: https://news.ycombinator.com/item?id=34616550 https://news.ycombinator.com/item?id=34616550 https://news.ycombinator.com/item?id=35451704 https://news.ycombinator.com/item?id=35451704

Frummy 4y ago
I don't find it strange that paintings match weather patterns with statistical significance. Painters are directly inspired by what's in front of them, I find it more than reasonable that they snapshot the state of pollution in the atmosphere. If you view the trends at https://www.pnas.org/doi/10.1073/pnas.2219118120 https://www.pnas.org/doi/10.1073/pnas.2219118120 slightly under "Trends in Contrast in Paintings by Turner, Monet, and Others" there are made convincing statistical arguments and the search for true "cause and effect" on wider phenomena such as society and nature is historically difficult without chains of statistical reasoning. It's not like saying person a shot person b causing c which is clearcut, more like events in set a slightly overlap events in set b and I find that a reasonable argument although not axiomatic which the world really can't be very often. This is where the humanities aid science
- replygirl 4y ago
  the study attempts to use science to aid the humanities, and it's the humanities whose standards are not met here. the study makes a claim about recent history where there's a lot of record to draw on; the standard is to get confirmation in writing. i've had papers docked for the same thing even when the person grading it agreed the claim seemed obvious

Frummy 4y ago

I don't find it strange that paintings match weather patterns with statistical significance. Painters are directly inspired by what's in front of them, I find it more than reasonable that they snapshot the state of pollution in the atmosphere. If you view the trends at https://www.pnas.org/doi/10.1073/pnas.2219118120 https://www.pnas.org/doi/10.1073/pnas.2219118120 slightly under "Trends in Contrast in Paintings by Turner, Monet, and Others" there are made convincing statistical arguments and the search for true "cause and effect" on wider phenomena such as society and nature is historically difficult without chains of statistical reasoning. It's not like saying person a shot person b causing c which is clearcut, more like events in set a slightly overlap events in set b and I find that a reasonable argument although not axiomatic which the world really can't be very often. This is where the humanities aid science

replygirl 4y ago
the study attempts to use science to aid the humanities, and it's the humanities whose standards are not met here. the study makes a claim about recent history where there's a lot of record to draw on; the standard is to get confirmation in writing. i've had papers docked for the same thing even when the person grading it agreed the claim seemed obvious