Thursday, May 13, 2010

Answer: Searching for something someone once said

Well... as you can see from the comments, there are actually at least three different articles that quote me in the NY Times. (Nice job, David!)  

How should one approach a search question like this?   Answer:  Once you realize that the name "Dan Russell" is very, very common (what where my parents thinking in giving me such a high frequency name?), you'll need to start the search with a little care because there are bound to be lots of false hits.  

You can see this right away from the obvious query:  [ Dan Russell ]   There are just too many "other" (and obviously incorrect!) Dan Russells in the news.  

So the first realization is this:  

People often have a "professional" name that they use.  That is, scholars (and people who are slightly savvy about marketing) like to have a single nom de professional that is the same everywhere.  That way, when someone does a search for the papers and articles you've written, they all show up in the same place in the index.  

This practice pre-dates the web by a lot.  Authors often want all their publications to be found at the same place in the library stacks and in the same location in the accumulated index of articles.  Does anyone else remember the "Reader's Guide to Periodical Literature"  published by the H. W. Wilson company?  It was the way you found articles pre-web-search.  It was a giant set of books that indexes magazines and journals by author's name and article subject.  To use the Reader's Guide, you'd look up "Russell, Daniel M." to get to all of my articles.

(And conversely, when authors do NOT want all their books or articles together in one bucket, they'd adopt a nom de plume to keep their various identities separate.  This was widely considered a good thing when you'd write serious works and juvenalia or sexy works that you didn't want to taint your professional reputation.) 

So...  when searching for someone who, like me, writes a fair bit and writes professionally--
try searching their professional / canonical name.  

The search [ Daniel M Russell ] is a bit better, but this is a case where you really want to limit by using site:  

[ Daniel M Russell ]  -- this starts to look much better.  

It's also worth knowing the conventions of the authors.  In this case, it's useful to know that the NYTimes has a policy of spelling out the names of people in their stories.  (Thus, looking for "Dan Russell' or "D. Russell" won't usually work... in the NYTimes... unless that IS the person's name, such as singer "Jay-Z," whose name actually includes the hyphen.)  

Now, having said all that, you should know that reporters also sometimes get it wrong, which means you still have to check name variants.  

So now where are we?  We have to try name variants ("Dan Russell"  "Daniel M Russell" and "Daniel Russell") and we'll use  But if you've been trying these queries out, you'll see that my name is STILL too common.  

We need to add something else to the query to cut down on all of the other, obviously bogus Dan Russells in the results!  

The simplest tactic at this point is to ADD a term (a "context term") that would discriminate the Dan Russells from each other.  Often a good filter is the place of employment.  In my case, that history includes Google and IBM.  So let's try adding in "IBM" as a context term.  

If we try adding "IBM" then you'll see that results get much better.  Now, the top 4 hits are all exactly what we're looking for.  

[ Daniel Russell IBM ]   

And, as you'll see (and just as David said), I've appeared 4 times in the NYTimes.  

The take-aways here are simple:  

1.  Use the SITE: operator to limit your search to just the newspaper of interest. 

2.  Search on name variants.  (And use double quotes if you want to limit the synonyms that Google will search for.  Note that the searches [ "Dan Russell" ] is very different from [ Dan Russell ] Without the quotes, Google will also search for Daniel (but not, so far as I know, for Danny)

3.  Search for a person's professional name (if they have one--many academics, writers and reporters do).  

4.  Remember to search for earlier variants of a name.  Women (and on occasion, men) will sometimes change their surname upon marriage.  Often the best way to discover earlier variants of a name is to look for their home web page or for parenthetical comments in their writings.  Look for things like  Susan Smith (nee Kline) for clues to other names.  

5.  Look for a "context term" that is associated with the particular person you're searching for.  In the case of Dan Russell, that "context term" is IBM.  It's just another word that's used in association with the person you seek (and NOT associated with all the other people with the same name).   

Search on! 


  1. Now for a different strategy. I started with Daniel Russell, which didn't work. Then I did "Daniel * Russell", which was more encouraging, as I wasn't sure you used your middle name or initial. I noticed that a lot of the crap was death notices, so I added -late -death and got things that were closer. But I only found you when I changed to "Dan Russell", so your claim about the NYTimes using full names was wrong in this case. (I did use for all of them. )

  2. Well.. I *also* said that sometimes the reporters get it wrong! Which was the case here. Generally speaking, the Times prefers full-names. The reporter just thought my full-name was Dan Russell.

  3. You stated : "Once you realize that the name "Dan Russell" is very, very common (what where my parents thinking in giving me such a high frequency name?"

    Yes this would be a common name & would perhaps require you to add something to it if you wish to be found. It made me laugh though & think back at names of people I have met, which were much worse choices by parents. I once met a Frank Frank as well as a man named Ira Tate ( irritate came to mind when he said his name). So as you see some parents can be much more cruel.