RDA-L Some discoveries of search engine results

Posting to RDA-L

On 8/26/2015 10:54 PM, Amanda Cossham wrote:
‘Real research’ is a pretty elitist term that I don’t think is at all helpful. Of course we don’t want to exist in a filter bubble, but things slip through the cracks regardless, because there is no perfect system, no perfect (re)searcher, no perfect indexer or cataloguer, no perfect search. And the needs of a first year history student or someone investigating a land claim are in different places on a continuum that might have Nobel Prize winning research at the other end and a casual ‘What’s the latest gossip about Lorde’ at the other.

There seems a tension in discussions on this list between what is necessary or ideal in academic libraries and what is necessary or ideal in the rest of the library world.

These have been some very interesting points. My own take is that it is not so much a question of real/elite research vs. academic libraries vs. public libraries and so on, but something else. I realize that it may be distasteful to some, but–perhaps based on my own proletarian background–I have always viewed a library as much more of a machine, or a tool. A very complex machine to be sure, but as with all machines and tools, there are right ways and wrong ways to use them.

So, to get the most out of a collection, if I know how to search a library’s tools: the catalogs, the indexes, etc. searches can be considered as “correct” or “incorrect”. And this works for all libraries, from the largest research libraries to the smallest local ones; that is, IF they follow the same basic method.

And what is that method? Libraries have always arranged their collections based on various interpretations of the rule of “consistency”. This means that similar things are described in similar ways, and that similar things are placed together so that others can find them. This is done so that if I find one item on the causes of WWI, I should–in theory–find all items on the causes of WWI. (I stress: this is in theory and in practice there are many exceptions)

Different libraries from different times and in different parts of the world will do this in different ways: rules, languages, and even concepts vary from time to time and from place to place. As a consequence, in a library it is a searcher’s task to find those points of consistency, no matter what rules, languages or concepts happen to be employed within that specific collection.

For those who attempt to find those points of consistency (and of course, we must assume that all of the information experts who are building these tools behind the scenes are also faithfully following the “rules of consistency”), you can determine definitively that some searches are bad, stupid, fair, good, better, best, or absolutely brilliant. This is similar to different abilities of driving a car (someone like me or a Mario Andretti), or using a power saw (someone like me or a professional carpenter). I am sure that Andretti could drive any car better than I ever could and a carpenter could use any saw better than I can, but put me in a library with them and I am the expert.

Compare this to search engines. I’ve been using them extensively since they began to exist; I have been studying them the entire time and read tons of books on how to search them. I have discovered that while there are definitely some searches that can be labelled “stupid” or “bad” (e.g. I am interested in the causes of WWI but I make a typo and put in WWII or WWIII–but who knows? Even THAT might work!) I have no idea how any search could possibly be labelled good, better, best or brilliant. The reliance on “relevance” makes everything completely different. So, while I can get a result that may appear to be “reasonable”–to me–that’s about the best I can do. There is no principle of consistency to compare it with.

For all I know, Andretti and the carpenter may use search engines better than I do. It is impossible for anybody to know for sure.

“Library consistency” is much easier discussed than achieved for both librarians and searchers, and it can be very difficult to find those points of consistency, especially for those who are not experts and they need help. That is the critical role of the reference librarian who helps people find those points of consistency.

I will also suggest that although an organization may have a huge amount of information, then while that entity may be important, and it may be necessary–yet if it does not follow such a basic principle as consistency, it cannot be called a library.

Using different words, I described this on the Bibframe list sometime back, and someone said that they were not interested in history. I replied that this is not history but still the same way libraries are designed to work today. That has never changed. To talk about the reality today however, we must take away the expertise of the searcher, remove the aids to searching the catalog (our cross-references do not work in a keyword environment) and remove the reference librarian, and it shouldn’t be a surprise that everything breaks down. That’s why the public doesn’t like library tools: we have removed all of the aids to searching them. (That’s far from all the problems, but a big one)

And maintaining that the answer is that everybody in the world should forever take information literacy workshops, is simply living in a world that never did, and never can, exist. From the beginning of establishing those training sessions, people immediately forgot what they learned, and they complained constantly. When the simpler alternative showed up (the search engines) they very naturally turned to them.

At the same time, my experience has been when/if people really understand that there is a genuine power that comes with the rule of consistency, and they can find it nowhere else, they absolutely love it and immediately become less enamored of “relevance ranking”. But nevertheless, saying that users need to suffer through boring IL workshops is actually hiding the fact that our tools need to be made easier.

So to me, the answer is rather obvious: we should design tools that take away the tedium from the library research process. I think modern computers are powerful enough so that there are many, many ways to do it–but the first step is to look at everything not outward from the viewpoint of the individual librarian or cataloger, but from the users’ point of view where the library is only one point of a huge universe. Most users today are rushed, being pulled in 20 directions at once, and overwhelmed with deluges of information/misinformation/irrelevant information, all vying for their attention.