Different types of smartness

Smart search engine behaviour takes different forms. Today so far I’ve commented on two different threads on this forum, both of which relate to that smartness; one of them requests more or better smartness, essentially to the level of maintaining curated categories like “search only within furniture stores”, while the other complains that some already-implemented types of smartness are thwarting people’s searches by (if I may put it this way) refusing to believe that the person really intended to search for what they searched for, and giving substitute results for what they “must have meant”.

To me, there are several legitimate explanations for a lack of curated categories: not enough staff, not enough time, not enough money, or too difficult to do it properly, would each make sense.

But when “find pages containing this but not that” is regarded as a tricky case, … well, to me that’s odd, and even vaguely sinister. It seems obvious to me that since it’s simple to find all pages that contain both words, it’s equally simple to find all pages that contain one word but not the other.

What have I missed?

I guess I could somewhat jokingly say “How much good is having 2050s technology, if we’re still struggling to catch up to the 1950s?”

Linking this through to the response here:


suit / suite is a stemming issue, treating word stems as the same in order to allow for matching such as apple onto apples. There is a brief explanation of stemming in our Words In The Index video which might be useful.

I had hoped that making an explicit exception (by using the minus) would override stemming.

I would much rather face the frustration of getting no results because I put the wrong ending on a word, than the frustration of not being able to exclude an overwhelming amount of unwanted information. There is so much material out there. We are not struggling to find what we want; finding what we want is easy. The real trick is not finding what we don’t want.