Museum peregrinations: Silhouette and mannequins

The National Gallery of Victoria exhibition Drape features a Versace evening frock from 1996 posed next to a 1913 evening gown in the style of Paul Poiret's Directoire Revival. Viewers were meant to remark on the ways in which designers separated by decades have interpreted the ancient Greek chiton; but what I found more remarkable were the mannequins on which the dresses were displayed.



The Versace mannequin is noticeably taller, more statuesque and angular than the one displaying the Schmoiret, which is petite, with sloping shoulders and a swelling bosom. The 'contemporary' mannequin has a figure sculpted by diet and exercise, whereas the 'historical' mannequin's figure is sculpted by corsetry.

The catalogue entry for this garment noted that its innovation was to evoke an uncorseted figure while actually requiring a corset. The 1910s were a transitional period in silhouette, as the voluptuous, tightly corseted 'Gibson Girl' with her 'S-bend' figure gave way to the svelte, boyish 'flapper'. As well as Poiret, designers including Lucile (Lucy, Lady Duff Gordon) and Madeleine Vionnet were favouring less restrictive corsetry and a more naturalistic silhouette.

(As an aside, I really want to get my hands on a copy of Waist Not: The Migration Of The Waist 1800-1960 by Richard Martin. But I baulk at the price of international shipping.)

Meanwhile, the bias cut is often championed as flattering to the female body, but the Versace dress looks cruelly revealing: see how the pubescent, pointy breasts and hipbones jut through the dress, and even the stylised fat pad on the mannequin's stomach is visible. It reminds me of one of the 'origin stories' of the bra: improvised by New York socialite Mary Phelps Jacob, who in 1910 had bought a diaphanous evening gown but found it revealed her corset's whalebones.

At the Fashion and Textile Gallery in Sydney, Dreaming Of Dior author and curator Charlotte Smith told me that she prefers modern mannequins for displaying much of the Darnell Collection because these mannequins are quite slim, which suits the clothes. I should add here that Smith – the former wearer of some of these garments – is a tiny, birdlike woman.

A group of older dresses from the Darnell Collection (including the bustle dresses from 1884 and 1885, below, and a 1910 evening dress) are actually posed on child-sized mannequins with padded busts, hips and replica bustles. (Smith has the real period undergarments, but they're too fragile to use under the dresses. Besides, I think it's a shame to hide them that way.)





These images don't give the best idea of how small these garments are, but they are clearly clothes for adult women, so perhaps they do convey how weird it felt to imagine the bra-wearing kiddie mannequins beneath.

Smith's take on the small size is that women in the 19th century weren't as active and athletic as women nowadays, and also that people then tended to settle in the same areas where they'd grown up, so they lacked the 'hybrid vigour' of intermarriage that would result in taller offspring.

I'm still skeptical that either of these ideas completely explains the radical smallness of old clothes, but I am getting a sense for the ways in which any genealogy of clothing size depends on a concomitant genealogy of silhouette… and corsetry. The NGV measures its garments by "centre back length" and "waist, flat", which sometimes gives an indication of fit but sometimes doesn't. For instance, the Versace dress has a waist measurement of 32.5cm, whereas a gold lamé dress by Vionnet, c1927, has a waist measurement of 61.4cm. This doesn't mean the Vionnet was made for a larger woman, but rather that 1920s dresses had no structured waist.

While it's useful to think about mannequins in this case because they are judiciously chosen to show off particular garments in a 'museum-appropriate' way, it's a mistake to extrapolate that the sizes and shapes of retail mannequins reveal very much about the size and fit of clothing. Not only are the mannequins themselves abstracted for visual impact, but the clothing can also be carefully pinned by visual merchandisers, in the back where customers won't see it.

Years ago I wrote a feature on mannequins for The Age. It was originally commissioned by Sunday Life, whose then editor wanted me to write, basically, "Oooh, look, mannequins are getting larger to reflect 'real women'." This wasn't the picture I got from my own observations, nor from interviewing a mannequin maker. Also, I think Sunday Life was a little freaked out that I brought RealDolls into it. So they spiked it, but I sold it to the regular paper instead (albeit without the RealDoll stuff).

However, another bit of my research that never ended up making it into the final article was an interview I did with Daisy Veitch of Sharp Dummies in Adelaide. They are a really fascinating firm because they don't make retail display mannequins; rather, their dummies are meant to be used during the manufacturing process.

They've harvested extensive anthropometric data and used it to build human forms that don't just reflect the shapes of actual bodies, but are also specially padded so that their 'flesh' has the same 'give' as a person's. This makes them invaluable when fitting things like lingerie and jeans, where it's important that the clothing doesn't dig in. Rather than use a (live) house model, your company could buy one of these dummies.

I wonder if I still have the interview transcript somewhere. I'd actually love to interview Daisy again because anthropometric research is deeply implicated in clothing sizing; the world's first anthropometric study was done in Sydney for Berlei in 1926-27, and still forms the basis of Australian clothing sizes.