I received a graphic in a presentation with equally weighted items in it. Their order did not match my expectation, so I started to wonder why that would be. Finally I figured out that I was reading them in text reading direction and that order did not match the order in which the text explained them. Therefore my question is:

In case a graphic has no special points of interest, do we tend to read images in text reading order, i.e. from top to bottom and left to right? What is the reason for it?

As a consequence

  • I would make the proposal to change the order of elements to match the sequence they were explained in the text
  • I might need to update some images in another project where they are used in context of a right to left language.

From my personal try with the following picture, I agree to this theory. I stared at the middle first, not finding anything useful, then starting from top left to bottom right. Side question: is this test ok or did I make a mistake designing the test?

Test image

  • 1
    @ThomasW I propose moving this to cogsci.stackexchange.com where you'll be more likely to get a scientifically backed answer. My gut reaction is that we do not read left to right or top to down. I suspect its more pattern recognition as your eyes (saccades) scan for patterns and words. It would also be a cultural assumption since not all cultures read top left to bottom right. But again, I'm just theorizing and think this deserves to be moved to CogSci.
    – Ryan
    Apr 28, 2014 at 15:49
  • 2
    This question is more about the theory of how we perceive things and should be moved to cogsci.stackexchange.com to get a more factually backed answer.
    – Ryan
    Apr 28, 2014 at 15:50
  • 1
    In this example my eyes are first drawn to the capitalized letters in the center/top. H then N particularly. They are larger and more visually interesting than an "I" or the series of surrounding "x" characters. I don't think there is a preset read direction, you are driving my eyes north westerly based on the appeal of what my eyes are drawn to, so after H and N, my view curls back around past the C then Y and down/around.
    – John
    Apr 28, 2014 at 17:44
  • Are you asking about one image (like a large photo) or a grid of individual images? I'm a bit confused as to what you are asking.
    – DA01
    Apr 28, 2014 at 17:51

5 Answers 5


I think the test is incorrect. These are not images but letters. It's normal for our minds to try and read the letters although they are separated in boxes. If they are actual pictures, I think we wouldn't "read" them in any particular order but instantly see the one that is most interesting to us. Size would also play part here. If the images are big enough to look at their detail, we would scan them in a more chaotic order than if they were small and lacking detail. I think the correct way to study this topic is to display a matrix of images for a brief time (1-2 seconds) and ask the participants to remember what they saw. If they remember the top left picture you can prove your point. My bet is they would remember images that are near the center. I also think that they would scan LtR if they are asked to remember as much as possible. They would try to make an array of images by remembering them LtR.


If the images are merely placed for no reason in the order they appear, then "yeah, of course" they should be rearranged by text order. Even if the arrangement is less pleasing aesthetically.

If there is a rational order to the images (time sequence, color arrangement etc,etc), then that order should take precedence and to hell with the text.

There is a huge difference between "expectation" and "imperative," and perception and understanding are not so rigid as to demand conformity to habit.


The answer on your question is picture/age/sex/task/etc-dependant: human first seeking signs of danger and on the other places all other.

The question was investigated using eye tracking equipment and the answer is probably "No" even when we exposed to text.

Look at the examples:

Site design related eye tracking -

enter image description here

Sex related eye tracking -

enter image description here

other example -

enter image description here


enter image description here

As you can see there is no line-related pattern of gaze. The gaze "takes" the picture more by slices, then by lines even when we read text and books.

wiki example -

enter image description here

  • 1
    Your examples are pictures with a single scene, not with several equally weighted items. What happened if we had an array of 20 photos of white, blonde naked top models taken in a similar environment? Would we start going through them in LtR order? Apr 29, 2014 at 18:26
  • @ThomasW. I don't believe in any way the human will follow any direction (remember, there are a lot of people reading from right to left...
    – Ilan
    Apr 29, 2014 at 18:29
  • Ilan, that's exactly what my question is about. Would Western people prefer "reading" pictures LtR while Eastern people do it the opposite way round? Apr 29, 2014 at 18:34
  • @ThomasW. Probably, yes. "In the same way, we all have a dominant or preferred “direction of processing” although the degree of dominance varies from person to person. When walking through a forest, most people who are right-handed will tend to brush the branches aside sweeping their right hand from left to right. (vice versa for people who are left-handed). However, not all right-handed people will have a dominant direction of processing from left to right (vice versa for left-handed people)" dyslexia-at-bay.com/…
    – Ilan
    Apr 29, 2014 at 18:46

In general I would say no, you do not perceive images in text reading order.

Images, unlike text, are heavily weighted by many factors - color, depth, motion, etc. These factors can pull the eye towards any image placed anywhere on the page. Your test is using all similar images, no image is weighted more than another (ignoring actually reading the characters within) - therefore it's not really a valid test.

It has also been proven that the bottom left corner of a layout is the least visible location for information. With text, you are lead right to that particular corner to continue reading. With images, that corner can be a "dead zone" and regardless of what image is placed there may result in a lack of attention from the viewer merely due to its location.

The bottom right corner is considered a "terminal" area where the reader is expecting things to end, or items requiring attention to continue are placed. To this end, images in the bottom right corner are going to naturally gain more focus that other images. If those images also happen to be prominent due to color, motion, etc. they can heavily weight that corner, again regardless of any natural text flow direction.

A bright, red, image in the bottom, left, middle, corner when all the other images are blue will naturally yank your eye right to that red image. It has nothing to do with text direction.

  • Scott is spot on IMO. That is what makes the difference between art and design. You need to be mindful of the page (or image) elements and it is your job as designer to tell the viewer where to look first, then second etc. As design has a purpose other than to just "look pretty", your ultimate goal, often a call to action (button for example) needs to be one of the things drawing the initial focus from the viewer. This is also why neutral themes are popular in design. A white or mostly neutral page with a big green button gets you looking in the right place from the start.
    – John
    Apr 29, 2014 at 17:00
  • But isn't top left to bottom right exactly the text reading direction LtR, left to right? The linked website did testing with right-handed people, but not with people from Arabia (or any other country with RtL reading direction). So we don't know yet whether they have their terminal area at the bottom left. Apr 29, 2014 at 18:31
  • My answer was for LTR countries. The terminal area for RTL countries would be bottom left. With images things are still very highly dependent upon the images themselves.
    – Scott
    Apr 29, 2014 at 18:33
  • So then your answer would be "yes, we perceive images in text reading order". Maybe I misunderstood this and you meant In general no (bold) but in this specific case yes (a bit hidden inside the non-bold text). Apr 29, 2014 at 18:38
  • No. (Speaking LTR countries) If a bright red image is in the bottom left corner of a grid consisting of blue images, the red image takes preciseness regardless of the natural tendency for that corner to be a low visibility area. Reading direction is only a slight precursor to view expectations. But those expectations can easily be circumvented based on image contents.
    – Scott
    Apr 29, 2014 at 18:42

Equally-weighted Items in a Grid

Based on your test image, I'm assuming it's a graphic in which the items were laid out in a grid format. With grid formats, the example that immediately comes to mind is that of comics/graphic novels, which are designed with the assumption that people will follow the action in a text reading direction.

Equally-weighted Items NOT in a Grid

Assuming that equally-weighted just means that the items are all roughly the same size and visually equally impactful, it's possible to have the items laid out in a way that directs your attention to the next item you should be looking at. For example, the items could be laid out in a regular circle to encourage a clockwise perusal method.

Side Answer

The test uses letters, which might bias people towards using text reading from the beginning. As well, some of the letters are lowercase, which makes them smaller. Perhaps you could use geometric shapes instead?

Other Stuff

From a graphic design perspective, Things Organised Neatly is a good place to look at graphics with many items to explore what layouts might work with your items.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.