Setting up new research project

DEVONthink Tips
1

I’m running DTPO (v. 1.5) under Leopard 10.5.2, and am relatively new to all this. So I’d like to ask advice about how to best set up a research project.

I’m considering writing about a famous person, let’s say for the sake of this example, it’s President Hoover. (It’s not, of course, but I’m superstitious about talking about subjects.) I’ve got half a dozen books written about him that I’m going to scan. I’ll create groups to cover the periods of his life (i.e., Early Childhood, Teen Years, College, Military Service, Early political work, Running for president, Presidency, Post-presidency, etc.) and probably sub-groups under those categories for specific events.

So, for example, under college years, I’ll have the six different versions of his college life in the one group so I can compare and amalgamate.

First question: what’s the most efficient way to get all this scanned material into the various groups? Copy and paste? Link?

Second question: I’ll have a second set of groups for non-event subjects, such as personal qualities. So, if one of the anecdotes from his college years also pertains to “Honesty”, I’d want the same text to appear in both groups. What’s the best way to do that?

Third question: is there a file size limit I should be aware of? Am I better off setting documents in regular text files and linking to them from DT, or is it okay to just load up the DT database?

And finally, is there a good tutorial or example somewhere on using DT for a large research project?

Thanks in advance…

/rb

2

One more question:

If I’m putting different parts of one book into different Groups, I will someday come across a paragraph in a certain folder and wonder where it came from. Can I automatically “tag” each selection with the name of the source and page number? Or do I have to manually tag each item?

3

There’s probably no single nor simple answer, depending an the subject matter, the type of project, and your personal preferences.

When I start a project, I create a new group for it. Inside that group I usually start with a new rich text note that begins with a summary of the topic and my objectives. Then the structure of the article or report is laid out, usually in the form of section and subsection headings.

For me, this is a stage that is important. As I define the structure I’ll often make inserted notes about each component including central ideas. This is a living summary and outline of the project. In the course of the actual writing I will probably revisit and modify these notes several times. I find that this helps me clarify the job at hand and stay on target.

Next, I’ll create subgroups corresponding to the principle structure headers, and of course in each a new rich text note with the corresponding name.

Now for the reference materials. In your case you are scanning and OCRing books and excerpts of booKs into your database. You likely have other materials as well, such as journal articles, news articles and your own notes.

I will create one or more subgroups for reference materials such as primary, secondary and perhaps for excerpts and quotations.

If I don’t plan to modify a document I will replicate it into a project reference subgroup. But if I may annotate a database document, or hack it into pieces, I will duplicate it into the project reference collection, leaving the original copy safe. If I want to break up a long PDF into chapters or even smaller excerpts, it’s easier to work with a converted rich text version of it. Of course, if you are making excerpts, make certain there’s information to document the source and page number.

I like to work inside my database, because I find it a rich working environment. But I have a friend who would still tackle a project like yours with stacks of 3 x 5 cards. As a former governor of Louisiana said, “There are more ways of killing a cat than by stuffing it with butter.”

4

This is one of those points where DT doesn’t make things as smooth as some of us might wish. Although in principle there’s nothing difficult about adding source info by copy and paste, it’s easy from time to time to forget to do this, causing lots of extra work later on. No doubt it’s easier to dream up some kind of automated tracking of where text has come from than to actually implement it, though.