Hi All,
I've found this board very very helpful but after searching through its archives I couldn't find answers to two pressing questions. Perhaps you could help me.
Background: I'm attempting to test hypotheses about the reaction in the 1750s to the Lisbon Earthquake of 1755. Contrary to the dominant interpretation of the event as causing great cognitive changes to the 'European mind', I find that most people responded with fear and interpreted the event as God's punishment. I have sampled sermons from before and after the event (in English). I now need tech assistance to complete the second and third parts of my project.
Second: extract 100 characters before and after each usage of 'God' and 'earthquake'
Third: using resulting segments of text, test whether the preponderance of terms in those samples cue *fear* as opposed to any other emotion
Fourth: I will be hand coding these segments myself for emotional content, and I may hire mTurkers to also code for me.
Questions: (1) RE step 2, does happen to know of anything that can automatically sample documents by removing a certain number of characters or words before and after a target term?
(2) Does anyone happen to know whether LIWC [http://liwc.net/howliwcworks.php] or something else would offer me the most explanatory power accounting for the emotional profiles of these documents? It seems LIWC is the DH standard for this sort of thing. But I wonder about its utility for analysis of historical documents. I suppose I would need to create my own dictionary of terms for 'fear', or perhaps add historical 18th terms to the native LIWC psychometric dictionary category of 'fear'? Maybe I'm way off. I haven't done this before.
