Nice article BTW

I completely agree with where the article landed, and found myself getting anxious along the way. That visual hierarchy of “things” rather than site structure or path is spot on.

I wonder if there has been any recent work regarding use of the back button as a means of navigation? How many people use it? How many people trust it? There have been several usability tests that I have observed that demonstrated to me how some folks just don’t use even their basic browser features. This might be something that changes over time, and that was a while ago.

Anecdotally, I recall navigating the Amazon breadcrumb (categories) while I was shopping for Christmas gifts for kids, and liking it quite a bit. It was really a means of modifying my search criteria. It was there, so I used it, but it could have been done in a lot of other ways and been even effective for me. The way that EUI handled search criteria in the eBay desktop app, or the way the Mac Mail app handles search. It’s a search/browse hybrid approach that works really well, and feels a bit breadcrumby – though really it’s not a breadcrumb at all…

Maybe context is key. It usually is. Some examples here are almost solely info-gathering situations. Shopping might be different and need some other model? I would love to see this idea systematically applied to a few contexts. You know, separate the apples from the oranges and see if the premise still holds as strong as it seems here.

@bobby By dynamic i mean that the pages are assembled by the server as and when the user requests rather than served as files off a file system. But as soon as they’re assembled the web takes over to distribute so again it’s HTTP and it’s stateless.

@xanti Erm, thanks for your thoughts. Quality of expression still counts…

@david Pre-web hypertext systems worked pretty much like that so all links were bidirectional. I think it’s interesting but damned tricky in real life. Backlinks are useful if the page isn’t heavily referenced but imagine a heavily linked to wikipedia page listing all the inbound links? It’s gonna get heavy quickly if it’s every linking wikipedia article, blog post, tweet…

@jakob Thanks and, erm, sorry for mispelling. I seem to have got it right for the first few paragraphs and got worse the more i typed. Is a wonder you didn’t end up as Jack :-)
I take your point on the reflect users’ mental models stuff but i think that’s what i was trying to say by:
“Many of the early battles in Information Architecture were about warping the filing systems and hierarchies that made sense inside businesses into filing systems and hierarchies that made sense to users”
Not saying those battles are won by any stretch but I think we have to move on from just content and page categorisation and think about the concepts those pages describe and the structure of the links between those concepts. Not sure attempts to force the web into a filing system type metaphor ever really work…

@debbie not much to say really > “my general feeling is that there is so much overlap in the fields of psychology and the web (possibly unsurprisingly) and that we could all benefit hugely from merging the two more” < totally agree :-)

As someone who is involved in the web but whose background is in psychology, the main thing that struck me about this post and what I really wanted to comment on is that in some ways what you have stumbled upon and are dealing with in breadcrumb trails is a debate that has raged in psychological circles for years – namely, how do people organise information?

If breadcrumbs are designed to prevent the user from feeling ‘lost’ or to signpost their location, then in some ways they should be fundamentally built around the psychology of the user.

I agree with you that path based trails are often meaningless and wholeheartedly agree with your point that “it’s not really about where you’ve come from” – I think these kinds of trails just lead to more confusion.

I also agree with the multi-faceted points you make about location based trails and the difficulty in classifying categories and hierarchies within these trails. Unfortunately however, this is not a new debate nor one that is confined to the web. It is one that psychologists have been researching for decades. How do we learn different categories and how do we place new objects into categories? Unfortunately, there are no easy answers to this debate, either in psychology or on the web. Different schools of psychological theory have yet to agree on how we form representations of knowledge (Do we use images or words? Hierarchies or more abstract structures? And so on) and have yet to agree on how we mentally categorise such knowledge (Do we form categories based on prototypes? Exemplars? Based on similarities between objects? Based on theories?)

This kind of debate is what underpins many of the difficulties presented by breadcrumb trails. The breadcrumb trail is trying to represent the location of a particular page in a way that is meaningful to the user, but even categorisation and classification from a user perspective is not simple.

I like your ideas about ‘thing structure’ rather than ‘site structure’ as I think these ideas represent categorisation and classification much more from a psychological perspective. I also like the ideas on the Wildlife Finder of using multiple forms of classification as this begins to represent more of the complexity found in categorisation. However, I fear that in many instances this could lead to overly complex ‘trails’ and more confusion rather than less, so I’m not sure I would advocate this in too many instances.

Overall, my general feeling is that there is so much overlap in the fields of psychology and the web (possibly unsurprisingly) and that we could all benefit hugely from merging the two more. Although I obviously haven’t really come up with any ‘answers’ here, I wonder if engaging in the ‘categorisation’ debate had in psychological circles could inform some of the problems found in online navigation.

(If you are interested, I have also written a blog in response to your article that provides a little more detail as this comment was already getting rather lengthy! (http://www.searchtosales.co.uk/blog/the-psychology-of-breadcrumb-trails/)