Applying the Gestalt Theory of Design to the 1993 Citroen Xantia

The A-pillar of the 1993 Citroen Xantia has always looked wrong. Now I know why and the reason it had to look that way. 

1993 Citroen Xantia as designed by Velizy.
1993 Citroen Xantia as designed by Velizy.

Unhappy with the actual car, I decided to see what would happen if I reduced the extent of the mirror sail panel so that more of the A-pillar showed. Read on to see how it looks if modified a little bit. It’s not as simple as just changing the shape of the black plastic part though…

First, what’s wrong. As designed the 1993 Xantia looks like this, below. The A-pillar disappears behind the black plastic of the mirror sail panel.

1993 Citroen Xantia as designed.
1993 Citroen Xantia as designed.

You are supposed to “read” the break-up of the A-pillar and glass graphics as shown below. One’s eye is supposed to follow a line from the side glass, across the mirror sail panel and then over to the trailing edge of the bonnet.

How you are supposed to see the break-up of glass, paint and plastic or dark and light colour fields.
How you are supposed to see the break-up of glass, paint and plastic or dark and light colour fields.

Only recently have I realised that that was how I was supposed to see it. The way I normally perceive the break-up of light and dark is like this:

1993 Citroen Xantia flow diagram-3

As a result it looks like this to my eyes:

1993 Citroen Xantia with red arrows attacking its a-pillar.
1993 Citroen Xantia with red arrows attacking its a-pillar.

So, a simple first attempt at a remedy is to reduce the extent of the mirror sail panel and align it better to the sideglass as shown here:

1993 Citroen Xantia, modified a-pillar.
1993 Citroen Xantia, modified a-pillar.

These two photos show the same thing, before and after:

1993 Citroen Xantia as designed.
1993 Citroen Xantia as designed.
1993 Citroen Xantai revised.
1993 Citroen Xantia revised. I don´t like the vertical line of the door´s leading edge. And the car looks more like a Peugeot 406 now.

The re-design looked better from some angles but from the front three-quarter still looked incorrect. To completely solve the problem would mean a different front-door/front-wing/A-pillar shutline. Citroen painted themselves into a corner here. And now I understand why Citroen couldn’t change that shutline. It would have meant a huge expense in modifying all the associated parts shared with the 406.

...and indeed it does. Here´s a Peugeot 406.
…and indeed it does.
Here´s a Peugeot 406.

On reflection, another solution is to handle the A-pillar the way they did it on the XM. That would mean blacking out the A-pillar and having a “cantilevered” roof graphic. Citroen probably didn’t attempt that because the XM was by 1993 bombing massively in the market.

What have we learned? First, that Gestalt theory explains why the figure-ground relationships of light and dark were not read as intended. In Gestalt theory, edges belong to the figure and not the field or ground it is sitting in or on. A simple example is the rub strip on the side of the car. In Gestalt theory the edge belongs to it and not the painted area around it.


We see the rub-strip as a figure on a ground of white paint in the above photos. In a more complex case, the black plastic of the mirror sail and the black or dark edges of the glass are seen as figures on a pale ground (on lighter coloured cars anyway). The edges belong to them.

One’s eyes see the mirror sail and side glass as one field, the windscreen as another and they are divided by the white or light A-pillar which terminates in a point. If Citroen’s designers had been more appraised of Gestalt theory and perhaps, more saliently, allowed more money they could have avoided a graphical failure like this.

Second, we’ve learned why Citroen had to do this, to make the car look less like the 406 with which it shared important underlying elements.

Author: richard herriott

I like anchovies. I dislike post-war town planning.

13 thoughts on “Applying the Gestalt Theory of Design to the 1993 Citroen Xantia”

  1. Bertone designed the Daewoo Espero of 1990. The most significant design element is the same mirror panel and A-pillar treatment, which is the worst element of both the 1993 Xantia and 1990 Espero. Is it possible that Citroen signed off on the Xantia just as the Espero was revealed – or did they really agree to give money to Bertone for a design already in production? I don’t imagine Bertone ever got another PSA contract. Since car designers have far more ideas than can be realised you’d imagine Bertone might have had at least one other sketch to hand that wasn’t 70% Daewoo.

  2. I agree the pillar treatment is ‘wrong’ though, for me, I’d analyse it as the fact that the A pillar appears to have no support. It looks functionally weak. Here’s another possible remedy.

    PSA have a history of taking over designs used or rejected by others, with Bertone’s earlier BX (already touted round to Volvo and Reliant) and the Pininfarina’s 605 (doubled up with Alfa’s 164). My current view of many of PSA’s woes is that Peugeot are excessively tight fisted. That isn’t a characteristic I find attractive in people, in a car manufacturer it’s potentially suicidal. Maybe the Italians gave them a generous discount for re-cycling.

    1. I think your analysis of the functional weakness is the same as mine. The pillar appears to rest on a pointy feature. It doesn´t look planted. This is what the image with the red arrows attacking the car is trying to get at.

  3. Interesting.
    I wonder if there is a difference in perception between pictures and real life (2D vs. 3D?). Is Gestalt theory actually applicable to pictures? And then, I have the feeling (one that I cannot substantiate with empirical facts) that cars are more and more designed for pictures and less for the road.
    That being said, I always liked the Xantia design-wise, the X2 even more so.

    1. Gestalt theory applies to any colour fields, two dimension or three. It might be that what looks right in 3D fails in 2D or fails in both, works in both or fails in 2D but not three D. Gestalt is just a theory about how people view colour fields. I know some people do like the Xantia; I don´t see how they miss the bad A-pillar though.

  4. Seant: your solution looks nice but…it involves a different style of door frame entirely which is why they were stuck with it for both cars. Yours has the window frame sitting into recess and the A-pillar bodycolour part stays fixed. In the PSA solution the doorframe is covering the actual A-pillar. I think Fiat did this with the Punto 2 as well. I prefer your solution, it´s neat and tidy though.

    1. No, I was rather assuming that, with a bit of thoughtful engineering, the ‘A pillar’ would hinge round with the door. Like Daniel, although I accept your A pillar criticism entirely, I’ve always liked the Xantia for its proportions and, in particular, the fact that it looks good when viewed directly from above. This isn’t really noticed by many people, so is hardly a selling point, but good to see all the same.

  5. I had not noticed the degree of commonality between Xantia and 406 until you pointed it out. In that respect the two designs function quite well.

  6. That’s true, it’s not obvious. The ruse might have been more excellent if the Citroen had not been saddled with the dodgy design. Also – and this is personal to me – it’s a car I never accepted as a Citroen. Nobody else agrees with me on this so I won’t labour the point.

  7. Well, although liking the Xantia I see your point.
    But on the other hand: What is a real Citroen below the upper middle class (that is: excluding C6 and XM) since Peugeot took over?

  8. Well quite. Up until the ZX they made sense and from then on they got less and less Citroeny, except for the C6, perhaps. At the time (circa 1990 something) this trend wasn’t clear to me. Now, two decades in I expect nothing from them other than that they don’t explode.

  9. I do think this A pillar shape is actually very smart and elegant, I think you and many other are looking at it in the wrong way and want to see something that’s PURPOSELY not there. This design piece is undoubtedly something people can’t notice at first glance, and I don’t think it was the designers’ intention.

    I think the goal was to propose an ingenious and subtle solution at connecting the fenders to the A pillar, in a way that’s not a smooth transition that’d break the pointy/sharp shape or overall design goal of the car, nor be an awkward connection where both sits against each other without any logical flow in the line or connection (the 406 sedan for example has a sharp connection between these elements, they don’t flow against each other, they just are there against each other, as opposed to the 406 coupe that has its hood smoothly joining the A pillar, which is a totally different design idea than the 406 sedan and overall look, (Image showing what I mean: This shows very well then these kind of pillars connection produce very different overall result) .
    It’s also interesting you mention the XM and how it did better, when the positioning of elements were the exact same as the Xantia, they just hid the A pillar and thus didn’t need this specific shape of the xantia, but the window, mirror and fender sit against each other exactly like in the Xantia. There to see it: (note that the red line on the door is the exact shape of the upper part of the door, it’s not just placed there to look like each other, it’s the line they both visually draw).

    The genius behind this A pillar idea is that it keeps a sharp connecting point. It is when you look at the bigger picture of what is happening that you realize the overall line of BOTH, the upper part of the fender and hood is thrown subtly and smoothly back toward the A pillar: . The thin “weak” part of the A pillar is exactly what makes the design of the xantia work as it is while keeping a logical connection between these elements At first glance, the A pillar doesn’t connect with the fender, and it very much WAS NEEDED for the sharp design the xantia had. If you check for the 406 coupe, or the C5 I First generations, this kind of smooth shape requires a quite rounded car, completely opposite of what the xantia came up with in 93, and for proof, nowadays where we’re going back at more squared /sharp cars, most cars today have gone back to a sharp A pillar, the C5X has a floating A pillar with no proper connection, the c5 II has a sharp A pillar connection, the bmw g30 has a sharp connection too. Hence once again, why this design on the Xantia is a brillant idea of playing with the contrast of elements to create a connecting illusion. It also allows for the windows to sit much lower and blend in with the mirrors, which again mirrors aren’t the most easy thing to integrate on a car. But low windows allows to increase the “pointy” feeling of the car as the rear will be able to feel like it sits much higher, when it reality it isn’t that high, and also better visibility…

    I understand the way you’ve been looking at it though, you’ve been expecting a design where the overall windows were separated into its own place, where they make a design of themselves, and you thus instinctively tried to look at the shape the windows were making, and looked around its outlines, and it didn’t make sense obviously. But that’s why it is brillant in my opinion, it’s meant to be looked with the car !

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: