How do you roll? Untangle the Great Buttondown Collar Conspiracy!

[Big A]

All the pictures look like anatomy diagrams to me for some reason

 

[ord65]

Jeezus--I don't want to sound patronizing or condescending, but you guys are capable of interpreting an illustration as a representation of reality, aren't you? You're tripping all over yourselves.

The illustrations represent worn collars.

I am an architect, sitting in an architect's office, drawing illustrations that represents (future) reality, surrounded by drawings that represent reality. I get it. I just didn't understand why you wouldn't judge a folded shirt - couldn't it be illustrative of how a worn shirt could look? Anyway, thanks for the thread - I've narrowed down what i like in a button down:
1. Relatively low amplitude roll
3. A stretched out bell shape that's verging on symmetrical
4. A nice roll fit around the knot (no billowing out)
5. No tie showing between the knot and collar
6. Moderately long points

 

[Manton]



All the pictures look like anatomy diagrams to me for some reason

maybe Georgia o'keefe paintings?

 

[mack11211]



Jeezus--I don't want to sound patronizing or condescending,

Too late!

The situation gets me thinking of that late point in DEAD RINGERS, when Jeremy Irons, playing one or the other of the twin gynecologists, complains that the patients don't fit the instruments.

 

[ord65]

anatomy, georgia O'keefe, gynecologists....I'll never wear another buttondown now.

 

[Trompe le Monde]



I think you are reading too much into this. .

 

[gimmeabreak]

For me the winner is # 1. I think it is accomplished through relatively long collar points, buttoned somewhat high (e.g. relative to #4), and somewhat spread, so the roll it pretty somewhat wide and flat on top (while preserving enough collar point length). I think it is ideal because this layout prevents two common problems with BD collars: looking too... pointy (e.g. not enough roll like in # 2), or having a roll such that they may collapse upon themselves (i suppose that's the exagerated S shape, and #4 and #5 look like they could be prone to this problem [problem for me anyway]).

 

[dieworkwear]

Not Brooksy whatsoever, but I like it.

 

[TheFoo]

Those are #1s.

 

[RDiaz]

If those are 1's then my vote should probably have been #1 rather than #4.

I didn't identify the "flat plateau" on top of either diagram, so just whent with the bell shape, which I like most on #4. But I do like that "plateau" to appear.

 

[DocHolliday]

Don't care for that (judging from the second pic). Too much.

 

[FlyingMonkey]

As far as I am concered, everything that needed to be said on this subject was said by dexconstruct on the heavytweed blog entry: the OCBD is a semi-casual shirt style whose joy is in movement and asymmetry, the perfect curves (whatever those are) that may appear are always only ever temporary - if you see them once, you may not see them again, but they may be another equally perfect at another time. Of course the button placement, and the weight of the fabric and the construction of the collar matter and they can all be very wrong, but it seems pretty clear that there is a wheel here that we don't need to reinvent.

 

[dieworkwear]



As far as I am concered, everything that needed to be said on this subject was said by dexconstruct on the heavytweed blog entry: the OCBD is a semi-casual shirt style whose joy is in movement and asymmetry, the perfect curves (whatever those are) that may appear are always only ever temporary - if you see them once, you may not see them again, but they may be another equally perfect at another time. Of course the button placement, and the weight of the fabric and the construction of the collar matter and they can all be very wrong, but it seems pretty clear that there is a wheel here that we don't need to reinvent.

I think to be fair to Foo, which I haven't been in this thread, there's a range of collar shapes that are appealing, and things outside that range are generally unappealing. It's not so much that there's one and only one shape, as that would be absurd, but there's a range of shapes that's generally good.

I don't think setting up these diagrams, however, and having people vote on them reveals much about that range. Pictures would probably be better.

I like these. As you can see, there's a range, and IMO, they're all good.

Spoiler: Warning: Spoiler!

I dislike the crisp folded one posted earlier and ord's short one above (no offense to either poster)

 

[RDiaz]

I'm wearing a linen BD today. I think if this collar was unlined it would be close to my ideal. It has very, very thin fusing. Not sure what illustration in the OP would represent it best.

 

[TheFoo]



For me the winner is # 1. I think it is accomplished through relatively long collar points, buttoned somewhat high (e.g. relative to #4), and somewhat spread, so the roll it pretty somewhat wide and flat on top (while preserving enough collar point length). I think it is ideal because this layout prevents two common problems with BD collars: looking too... pointy (e.g. not enough roll like in # 2), or having a roll such that they may collapse upon themselves (i suppose that's the exagerated S shape, and #4 and #5 look like they could be prone to this problem [problem for me anyway]).

The spread of #1 is no different from #4. If you click through the different collars in my blog post, this may be more evident. The only collar with more spread is #3.

As far as I am concered, everything that needed to be said on this subject was said by dexconstruct on the heavytweed blog entry: the OCBD is a semi-casual shirt style whose joy is in movement and asymmetry, the perfect curves (whatever those are) that may appear are always only ever temporary - if you see them once, you may not see them again, but they may be another equally perfect at another time. Of course the button placement, and the weight of the fabric and the construction of the collar matter and they can all be very wrong, but it seems pretty clear that there is a wheel here that we don't need to reinvent.

Again, I implore you not to assume there isn't anything to discuss.

The Heavytweed article is an excellent history of the OCBD. It is not, however, an attentive aesthetic account. That clearly wasn't his goal. So, yes, the feel, history and context of a thing are crucial--I have argued that myself, time and time again. However, it is a gigantic mistake to assume clothing is only about those "soft" factors. A common misstep on this forum is to assume that one can leverage such soft factors (usually attitude, emotion, lifestyle, etc.) to look good, without attending to technical execution. Hence, you get a lot of feeling over substance, which inevitably leads to lots of egg on face. If you don't know how to recognize poor fit, sloppy workmanship, styling incoherencies, etc., no amount of "joy" will compensate.

I suggest being more holistic in your approach.

Anyway, if you pay attention to Heavytweed's visual examples, you'll see that a couple key points I've been making are true. First of all, these collars do tend to take on certain shapes. Often, a bell shape. If you are obsessing over whether one side is symmetrical with the other, you are obsessing over the wrong thing. I assumed everyone could make allowances for real-life variances in choosing their favorite of the illustrations I offered. Given that, most of the live photos he's posted toward the end of his article look like #1. All of Brooks Brothers's own illustrations show something like #1 or #4. If you've ever worn an OCBD before, you should know that these shapes don't change much through wear. Yes, the collar will shift about. I don't see why you'd assume otherwise from a still drawing, but not a still photograph. However, the fundamental shape tends to stay intact. Your #1 or #4 will not suddenly turn into #2. #3 is simply cut different from the others, so it is also not something that occurs by happenstance. I've had #5 collars myself. They just stay that way--good or bad.

Second, when people praise the appearance of an OCBD collar, they are often praising very disparate things, which are not merely different points on a spectrum of amorphousness, but intrinsically different types of rolls that depend on design. Take a look at these two different collar rolls Heavytweed posted:

They are absolutely, completely different (#2 on left, #1 on right). Yet, both are vintage Brooks Brothers OCBDS, and he himself attributes the difference to manufacture and design, not the randomness of wear:

"This pink bold stripe oxford cloth shirt from the same period also has that unmistakable open collar roll that has become such an easily recognizable part of the American style . . . [In contrast,] the bell shaped collar roll is perhaps the button-collar that is most well-known and most easily achieved by nearly every shirt maker."

I'm no novice on this stuff. I've tangled quite a bit with OCBDs and trying to get the collar I idealize. Moreover, other experienced members have gone through similar ordeals. So, I didn't mean to invite hundreds of comments aiming to enlighten on the free-style spirit of the OCBD. My purpose is to elevate the discussion beyond that.

I think to be fair to Foo, which I haven't been in this thread, there's a range of collar shapes that are appealing, and things outside that range are generally unappealing. It's not so much that there's one and only one shape, as that would be absurd, but there's a range of shapes that's generally good.

I don't think setting up these diagrams, however, and having people vote on them reveals much about that range. Pictures would probably be better.

The problem with photographs of live examples is that people easily judge the wrong things, taking factors into account that have nothing to do with the collar itself. Remember, that is part of my hypothesis: that very few people are actually judging the roll of the collar when they look at pictures and say "nice roll!" They are often enamored with other, less tangible things.

So, yes, this exercise calls for some minor ability to associate the Platonic with reality, but the benefit is that it focuses attention. Anyway, I think the diagrams are pretty darn specific. It's not like I gave you five different triangles to compare.

I'm not sure why you're concerned that a preference for a range versus a single ideal is a problem here. If there is such a range, it will show itself in the results. So far, all indications are that any such range is quite small.