Cooper Legal Research

Call 24/7 (905)737-9994

Legal Research Tips

LEGAL RESEARCH TIP: THE VALUE OF KEYWORD PROXIMITY SEARCHING


Too often, lawyers and novice legal researchers get bogged down in the following scenario: 
 
You enter a keyword search of legal cases, only to find that you must cycle through 853 results.  You then enter more keywords into your query in the hope of knocking down the list.  Maybe you were able to narrow the list down to 53 results, or perhaps even as little as ten results.
 
But then you're faced with another problem - long, interminable cases littered with your targeted keywords, forcing you to spend more time scrolling through The-Case-That-Never-Ends, in a fruitless search for valuable dicta that is relevant to your legal issue. 
 
Soon enough, your brain is clouded by all that needless skimming, and you're just about ready to throw your laptop (or throw out your back hoisting your desktop) against the wall.
 
Had you known there were other time-saving legal research tactics open to you, you might have saved your brain (and back) the trouble.  One such time-saving tactic is the proximity search, which entails that you specify in your search the number of words that separate two or more of your targeted keywords.
 
For instance, let's say that you wanted to narrow your case law search to just those paragraphs that discuss the issue of imputing income in spousal support cases.
 
Your strategy might be to specify the number of words that separate the keywords or keyword phrases "impute", "income", and "spousal support".
 
Now, there are a number of case law databases out there, some free (like Canlii) and some that are only available by yearly subscription, like Quicklaw or Westlaw eCarswell.  Each of these databases employ what is called Boolean Search, a series of special commands that restrict the parameters of your search.
 
As I am most familiar with the Westlaw search protocols, I will be using it in my example; however, almost all these databases allow for keyword proximity searching, and so, the same strategies and concepts apply, even if the particular form of the search command may vary from database to database.
 
For  keyword proximity search with the Westlaw database, you specify the proximity length with the command "w/n".  So, for instance, if you want to look for a proximity distance between the keywords "impute" and "spousal support" of four words, you would specify the search command as:
 
impute w/4 "spousal support"
 
And voila!  Every case you find, you can then scroll directly to the paragraph containing these words in the specified proximity, which will be highlighted in your search results. 
 
But wait - there are other issues you should consider before considering yourself an expert proximity navigator.  What if your specified proximity between these words is too restrictive, meaning that your search command has possibly excluded valuable dicta that would have been discovered had you expanded your specified proximity to nine words, or perhaps even to as many as 20 words? 
 
But as you extend the proximity specification, you then open up the possibility that your gold nugget of dicta is again hidden among a veritable trove of results, raising that brain-numbing needle in a haystack effect again.
 
Perhaps the specification of proximity between the keywords "impute" and "spousal support" is not so optimal.  After all, a court can and does engage in long discussions of the subject of imputing income before or after using the term "spousal support."  Maybe, then, it is more appropriate to enter the following command:
 
impute w/4 income AND "spousal support"
 
Here, you have surmised that the term "impute" will most likely show up in very close proximity to the term "income."  You may also surmise that the further away the specification between these two keywords, the more likely you are to get irrelevant results in which the term "impute" is not conceptually related to the term "income."  
 
In this latest search command example, notice that I have also included the keyword phrase "spousal support", though I did not specify any proximity distance for it.  In practice, this means that I will likley get results that speak to the issue of imputing income, and which will also cover the issue of spousal support.  On the other hand, you will likely find that in many cases, the specific discussion on imputing income relates to the section of a decision dealing with child support; hence, that dicta is not relevant to your search.
 
So, again, in order to ensure that all your targeted terms bunch together in close, highlighted proximity, you might specify the following search command:
 
impute w/4 income w/100 "'spousal support"
 
Notice that I considerably expanded the proximity of the phrase "'spousal support"  to within 100 words of the terms impute w/4 income.  With this example, it is safe and indeed advisable to widen the proximity here, since the command is far more specific and detailed.  Here is a command that restricts your search to discussions of imputing income, and is wide enough to capture most - though, still, not necessarily all - such discussions conceptually dealing with this issue in the context of spousal support. 
 
But let us suppose that your issue is very specific, and that you don't need to cycle through all the varied dicta that deals with the issue of imputing income in a spousal support context.  Let us suppose that your issue is fact-specific, dealing with the issue of imputing income in a situation where the spousal support recipient is being supported by a boyfriend or girlfriend.  Your search command may then be as follows:
 
impute w/4 income w/100 "spousal support" w/100 boyfriend OR girlfriend
 
One caveat here is that the more detailed your search command, the more you risk excluding dicta that might have shown up had you chosen other proximities and keywords.  So, your search strategy entails an essential balance between culling and capturing your targeted information. 
 
The key take-away point is that, in general, you want to simultaneously cull your results by using close proximities for only those keywords that normally appear conceptually close together in the body of case law (like with the keywords "impute" and "income"), while at the same time seeking to exhaustively capture all the culled results that  contain keywords not usually associated with one another (like with the keywords "'spousal support," "boyfriend," and "girlfriend").  And when exhaustively capturing those less commonly related words, you need to develop an instinct for how far out you specify their proximity to your cosely proximated batch before they cross the conceptual threshold for irrelevant results.
 
In the end, when faced with very complex, abstract, or highly particular legal issues and/or facts, you will realize that there is more than one optimal way to skin that proverbial legal research cat. Legal research is an art, not a science - but it is an essential tool for finding the supporting precedent you'll need to nail down the legal basis for your litigation arguments. 
 
And one final caveat here:  you want to make sure that your search command captures all forms of your targeted keywords.  A search command for the keyword "impute", for example, does necessarily capture all instances where the keyword "imputing" occurs. 
 
But that is a legal research tip and lesson for another discussion...

***************************************************************
James Cooper, LL.B., MFA provides professional legal research assistance to law firms and members of the public throughout the GTA (Toronto, Mississauga, Vaughan, Richmond Hill, Newmarket, Barrie) and the Province of Ontario.  He may be reached for a free initial phone consultation at (905) 737-9994.  Email inquiries may be directed to jcooper at TorontoLegalResearch.com
 
 

Web Hosting Companies