6 ms·

This guy didn't just make up the rule of thirds, symmetry, or spatial rhythm. Its not unreasonable to try to understand why some types of architecture might be

This guy didn't just make up the rule of thirds, symmetry, or spatial rhythm. Its not unreasonable to try to understand why some types of architecture might be preferred over others. Rather than dismiss an entire field of inquiry, maybe try to see why it exists and if there is anything you can learn from it. Like much aesthetic theory, a lot of these analyses are post-hoc, but this isnt really a problem; its better to ask "why do people like this" than to state "people should like this because..." And sure, like much aesthetic theory, it can be a pretext for sociopolitical critique, but don't let that mar the whole thing.

Houshalter 10y ago
But a lot of these rules are silly pseudoscience. I am strongly reminded of the history of people obsessed with fitting everything into the golden ratio. Because it's "god's ratio" and supposedly appears everywhere in nature. It turns out a lot of that was BS. And it certainly didn't make paintings that didn't fit the golden ratio bad or inferior. There isn't any evidence to support that any of these rules actually make the architecture "better", whatever that even means. Even if a survey found that people prefer one style to another, it's still a highly subjective personal preference. There can be people with different preferences. If people didn't like this style of houses, they wouldn't be buying or producing them.
- mseebach 10y ago
  I think the general idea is that following the rules doesn't make your architecture good, it's that breaking them (probably) makes it bad. But if you break them in the right way in the right place for the right reason, it can be awesome. I guess code style can be a useful analogy: you can write absolutely horrendous, unmaintainable, unreadable code that follows all the best style practices, and you can have awesome code that breaks them. But when writing code in your day-to-day work, following the best practices, only breaking them when they're irretrievably in your way, is undoubtedly the best advice for not-super human non-geniuses to write good code. Most of the mansions showcased are well within the range of having been salvageable had the architect run checkstyle on the drafts.
  - Houshalter 10y ago
    But my point is nothing about breaking these rules makes it "bad". At least OP didn't present any evidence that it does. Just that he, personally, doesn't like it. At least with code one could make an argument that a certain style increases readability and makes it easier to follow. Although even there I think it's a lot of subjective preference, and it doesn't really matter as long as you are consistent. With these buildings, changing the style to fit these arbitrary rules could make them worse. Like reducing the number of windows just so it fits into a certain style. Windows are nice and provide better natural lighting. Studies have found windows and natural lighting have a a very positive effect on mental health. Or forcing the house to be symmetrical. That could make the layout suboptimal. As opposed to if it was allowed to be built without that constraint. And so on for the other rules. Every rule adds more constraints, which requires sacrificing other aspects to satisfy.
    - thucydides 10y ago
      OK. Hopefully I'm not putting words in your mouth. You think the author's explanation of why McMansions look like garbage is too soft and squishy. Too "subjective." In reality, it's all just personal taste, and as we know, there's no accounting for taste. Serious question: do you think this house looks beautiful? http://65.media.tumblr.com/5cfa243ca7f2997cf794ed190c24c38b/tumblr_inline_obrdug3QsV1sppt0x_500.jpg http://65.media.tumblr.com/5cfa243ca7f2997cf794ed190c24c38b/... Does it look better, worse, or the same as this house? https://upload.wikimedia.org/wikipedia/commons/3/37/Thomas_Jefferson's_Monticello.JPG https://upload.wikimedia.org/wikipedia/commons/3/37/Thomas_J... Do you think a large majority of people would agree with your personal, subjective taste in this case?
      - erichocean 10y ago
        Is there really a lawn on TJ's front porch? That looks fun to mow…
      - lake99 10y ago
        I agree with everything laretluval and Houshalter have said about the topic, and since they haven't answered your question, I'll do it. 1. I don't think it's fair to compare the two houses. The person who built the second house seems to have spent a lot more money than the person who build the first. 2. The first one gives the appearance that it has many tiny rooms, which I don't like. 3. I don't like the styling of either house. They both have weird styling elements (I don't the technical terms).
- avn2109 10y ago
  >> "...highly subjective personal preference." I'm sorry to be the smug guy linking PG on HN, but this essay [0] directly engages your point here and IMHO pretty conclusively refutes it. This is the money quote: "As in any job, as you continue to design things, you'll get better at it. Your tastes will change. And, like anyone who gets better at their job, you'll know you're getting better. If so, your old tastes were not merely different, but worse. Poof goes the axiom that taste can't be wrong." [0] http://www.paulgraham.com/taste.html http://www.paulgraham.com/taste.html
  - Houshalter 10y ago
    If the people buying and living in the house prefer it, how can you tell them their taste is wrong? What does it even mean for taste to be "wrong"? What are the actual consequences of it? If you go to a different culture, that has different food, architecture, music, art, etc, would you tell them they are wrong about everything and need to change? Would you listen to them if they told you your tastes were wrong? There is nothing objective about taste. It's entirely determined by associations you make in the brain. The author has associated "McMansion" architecture with things he doesn't like. Maybe the cheap construction they are associated with, or the people who live in those houses, etc. But other people can have completely different associations.
- Noos 10y ago
  I don't think they are psuedoscience, but it might be that architecture is an unusual case. With other art forms, yes they use these rules, but there's also things like line, texture, color, and other rules that disguise or soften them. With architecture, I think they don't have much technique to do so, and the symmetry, golden ratio, etc, come through much stronger. Many of the examples he gave in his blog of good symmetry were actually fairly boring houses, and the problem with too much symmetry is a boring picture. I mean the rules work, but in functional forms like houses, you have less ability to make them visually interesting overall. The McMansions might be the same principle in opposite-there isn't enough to make the non-symmetrical, almost abstract design less chaotic while keeping the visual interest.
tlb 10y ago
He assumes that the design of dwellings should be determined by the aesthetic quality of their exterior front view. But the owners spend the most time in the interior, and second most in the back yard.
- gdwatson 10y ago
  I too wish he'd written about more than the façade, but this piece is only the first in a series. He goes on to talk about garages and columns, so perhaps he'll get to interiors.
- dpark 10y ago
  If you want to make the best use of the interior space, you don't carve up a half dozen secondary masses (using the article's term). You do this because you have bad taste and think it looks fancy/expensive. i.e. You think it looks like a proper mansion when in reality your 3000 sqft house should not be carved up into half a dozen independent chunks as if it's a palace built up over 4 centuries. Maxing out the use of the interior space means one huge box, which is what most "ultramodern" houses in cities are. It's not my aesthetic, but it least it makes sense, unlike putting tacky form over function in the mcmansion style.
- jasonlotito 10y ago
  No, he doesn't. He's merely talking about one aspect of the design of dwellings. There are other articles talking about other aspects.
jasode 10y ago
>, a lot of these analyses are post-hoc, but this isnt really a problem; It actually is the source of the problem because his rules end up being inconsistent. There are billionaire estates in the ultra expensive zipcodes of the Hamptons[1] (NY) and Atherton[2] (near Silicon Valley) that people would not call "McMansions" that violate the author's principles about "voids and masses". Put another way... as a thought experiment, suppose we had the AutoCAD drawings of every house from modest cabins to mega estates. We attempt to write a computer program that has a boolean function called IsMcMansion(AutoCAD_file) that returns either TRUE or FALSE using the blog author's criteria. We can attempt to tweak the logic of the program to analyze and weigh the geometries of the "masses" and calculate the proportion of "voids" but we'll never end up with a program that 100% correctly predicts what humans label as "McMansion" Why? Because, those rules don't work! E.g.... IsMcMansion(famous_novelist_house)==TRUE is determined to be wrong because us humans look at that house and say "that house isn't a McMansion because it was built in 1914 and has so much _character_" That's what we humans do. We post-hoc a lot of judgments with a veneer of authoritative-sounding logic. We actually arrive at "McMansion" by combining a bunch of other attributes like: lot size, building materials, builder, socioeconomic rank of buyers, year of construction, trendy styles, etc. E.g. IsMcMansion(Graceland)==TRUE because of the algorithm penalized its oversized portico. However, it is judged by humans to be misclassified because... Elvis once lived there. Hence nobody calls the Graceland Estate a McMansion. If the exact same house was built by a national builder like Toll Brothers[3] and the people living in it were social-ladder-climbing yuppies, we very well might call it a "McMansion." [1]https://www.google.com/search?q=mansions+hamptons&source=lnms&tbm=isch https://www.google.com/search?q=mansions+hamptons&source=lnm... [2]https://www.google.com/search?q=mansions+atherton&source=lnms&tbm=isch https://www.google.com/search?q=mansions+atherton&source=lnm... [3]https://www.tollbrothers.com/ https://www.tollbrothers.com/
- wpietri 10y ago
  I don't quite get your point here. Nobody is suggesting that the term McMansion is purely about architecture. There have surely been terrible buildings as long as there have been buildings. But this particular kind of terrible building is notable because of the socioeconomic factors that created so many of them. The point of a McMansion is essentially social. Once you get past the structural basics, the point of most architecture is social. Humans are social primates. Yes, of course, if you construct a program without regard to sociality it will not be very good at understanding social phenomena.

Houshalter 10y ago

But a lot of these rules are silly pseudoscience. I am strongly reminded of the history of people obsessed with fitting everything into the golden ratio. Because it's "god's ratio" and supposedly appears everywhere in nature. It turns out a lot of that was BS. And it certainly didn't make paintings that didn't fit the golden ratio bad or inferior. There isn't any evidence to support that any of these rules actually make the architecture "better", whatever that even means. Even if a survey found that people prefer one style to another, it's still a highly subjective personal preference. There can be people with different preferences. If people didn't like this style of houses, they wouldn't be buying or producing them.

mseebach 10y ago
I think the general idea is that following the rules doesn't make your architecture good, it's that breaking them (probably) makes it bad. But if you break them in the right way in the right place for the right reason, it can be awesome. I guess code style can be a useful analogy: you can write absolutely horrendous, unmaintainable, unreadable code that follows all the best style practices, and you can have awesome code that breaks them. But when writing code in your day-to-day work, following the best practices, only breaking them when they're irretrievably in your way, is undoubtedly the best advice for not-super human non-geniuses to write good code. Most of the mansions showcased are well within the range of having been salvageable had the architect run checkstyle on the drafts.
- Houshalter 10y ago
  But my point is nothing about breaking these rules makes it "bad". At least OP didn't present any evidence that it does. Just that he, personally, doesn't like it. At least with code one could make an argument that a certain style increases readability and makes it easier to follow. Although even there I think it's a lot of subjective preference, and it doesn't really matter as long as you are consistent. With these buildings, changing the style to fit these arbitrary rules could make them worse. Like reducing the number of windows just so it fits into a certain style. Windows are nice and provide better natural lighting. Studies have found windows and natural lighting have a a very positive effect on mental health. Or forcing the house to be symmetrical. That could make the layout suboptimal. As opposed to if it was allowed to be built without that constraint. And so on for the other rules. Every rule adds more constraints, which requires sacrificing other aspects to satisfy.
  - thucydides 10y ago
    OK. Hopefully I'm not putting words in your mouth. You think the author's explanation of why McMansions look like garbage is too soft and squishy. Too "subjective." In reality, it's all just personal taste, and as we know, there's no accounting for taste. Serious question: do you think this house looks beautiful? http://65.media.tumblr.com/5cfa243ca7f2997cf794ed190c24c38b/tumblr_inline_obrdug3QsV1sppt0x_500.jpg http://65.media.tumblr.com/5cfa243ca7f2997cf794ed190c24c38b/... Does it look better, worse, or the same as this house? https://upload.wikimedia.org/wikipedia/commons/3/37/Thomas_Jefferson's_Monticello.JPG https://upload.wikimedia.org/wikipedia/commons/3/37/Thomas_J... Do you think a large majority of people would agree with your personal, subjective taste in this case?
    - erichocean 10y ago
      Is there really a lawn on TJ's front porch? That looks fun to mow…
    - lake99 10y ago
      I agree with everything laretluval and Houshalter have said about the topic, and since they haven't answered your question, I'll do it. 1. I don't think it's fair to compare the two houses. The person who built the second house seems to have spent a lot more money than the person who build the first. 2. The first one gives the appearance that it has many tiny rooms, which I don't like. 3. I don't like the styling of either house. They both have weird styling elements (I don't the technical terms).
avn2109 10y ago
>> "...highly subjective personal preference." I'm sorry to be the smug guy linking PG on HN, but this essay [0] directly engages your point here and IMHO pretty conclusively refutes it. This is the money quote: "As in any job, as you continue to design things, you'll get better at it. Your tastes will change. And, like anyone who gets better at their job, you'll know you're getting better. If so, your old tastes were not merely different, but worse. Poof goes the axiom that taste can't be wrong." [0] http://www.paulgraham.com/taste.html http://www.paulgraham.com/taste.html
- Houshalter 10y ago
  If the people buying and living in the house prefer it, how can you tell them their taste is wrong? What does it even mean for taste to be "wrong"? What are the actual consequences of it? If you go to a different culture, that has different food, architecture, music, art, etc, would you tell them they are wrong about everything and need to change? Would you listen to them if they told you your tastes were wrong? There is nothing objective about taste. It's entirely determined by associations you make in the brain. The author has associated "McMansion" architecture with things he doesn't like. Maybe the cheap construction they are associated with, or the people who live in those houses, etc. But other people can have completely different associations.
Noos 10y ago
I don't think they are psuedoscience, but it might be that architecture is an unusual case. With other art forms, yes they use these rules, but there's also things like line, texture, color, and other rules that disguise or soften them. With architecture, I think they don't have much technique to do so, and the symmetry, golden ratio, etc, come through much stronger. Many of the examples he gave in his blog of good symmetry were actually fairly boring houses, and the problem with too much symmetry is a boring picture. I mean the rules work, but in functional forms like houses, you have less ability to make them visually interesting overall. The McMansions might be the same principle in opposite-there isn't enough to make the non-symmetrical, almost abstract design less chaotic while keeping the visual interest.

tlb 10y ago

He assumes that the design of dwellings should be determined by the aesthetic quality of their exterior front view. But the owners spend the most time in the interior, and second most in the back yard.

gdwatson 10y ago
I too wish he'd written about more than the façade, but this piece is only the first in a series. He goes on to talk about garages and columns, so perhaps he'll get to interiors.
dpark 10y ago
If you want to make the best use of the interior space, you don't carve up a half dozen secondary masses (using the article's term). You do this because you have bad taste and think it looks fancy/expensive. i.e. You think it looks like a proper mansion when in reality your 3000 sqft house should not be carved up into half a dozen independent chunks as if it's a palace built up over 4 centuries. Maxing out the use of the interior space means one huge box, which is what most "ultramodern" houses in cities are. It's not my aesthetic, but it least it makes sense, unlike putting tacky form over function in the mcmansion style.
jasonlotito 10y ago
No, he doesn't. He's merely talking about one aspect of the design of dwellings. There are other articles talking about other aspects.

jasode 10y ago

>, a lot of these analyses are post-hoc, but this isnt really a problem; It actually is the source of the problem because his rules end up being inconsistent. There are billionaire estates in the ultra expensive zipcodes of the Hamptons[1] (NY) and Atherton[2] (near Silicon Valley) that people would not call "McMansions" that violate the author's principles about "voids and masses". Put another way... as a thought experiment, suppose we had the AutoCAD drawings of every house from modest cabins to mega estates. We attempt to write a computer program that has a boolean function called IsMcMansion(AutoCAD_file) that returns either TRUE or FALSE using the blog author's criteria. We can attempt to tweak the logic of the program to analyze and weigh the geometries of the "masses" and calculate the proportion of "voids" but we'll never end up with a program that 100% correctly predicts what humans label as "McMansion" Why? Because, those rules don't work! E.g.... IsMcMansion(famous_novelist_house)==TRUE is determined to be wrong because us humans look at that house and say "that house isn't a McMansion because it was built in 1914 and has so much _character_" That's what we humans do. We post-hoc a lot of judgments with a veneer of authoritative-sounding logic. We actually arrive at "McMansion" by combining a bunch of other attributes like: lot size, building materials, builder, socioeconomic rank of buyers, year of construction, trendy styles, etc. E.g. IsMcMansion(Graceland)==TRUE because of the algorithm penalized its oversized portico. However, it is judged by humans to be misclassified because... Elvis once lived there. Hence nobody calls the Graceland Estate a McMansion. If the exact same house was built by a national builder like Toll Brothers[3] and the people living in it were social-ladder-climbing yuppies, we very well might call it a "McMansion." [1]https://www.google.com/search?q=mansions+hamptons&source=lnms&tbm=isch https://www.google.com/search?q=mansions+hamptons&source=lnm... [2]https://www.google.com/search?q=mansions+atherton&source=lnms&tbm=isch https://www.google.com/search?q=mansions+atherton&source=lnm... [3]https://www.tollbrothers.com/ https://www.tollbrothers.com/

wpietri 10y ago
I don't quite get your point here. Nobody is suggesting that the term McMansion is purely about architecture. There have surely been terrible buildings as long as there have been buildings. But this particular kind of terrible building is notable because of the socioeconomic factors that created so many of them. The point of a McMansion is essentially social. Once you get past the structural basics, the point of most architecture is social. Humans are social primates. Yes, of course, if you construct a program without regard to sociality it will not be very good at understanding social phenomena.