Quantcast
Channel: kyramnelson -
Viewing all articles
Browse latest Browse all 86

What Makes a Successful Query (according to science)

0
0

Once upon a time I was in a PhD program in linguistics. Which means I spent a lot of time doing a lot of analyses on different types of language. For one of my class projects I looked specifically at query letters. I identified two linguistic distinctions between query letters that were successful and query letters that were unsuccessful. I was pretty interested in what I found and wanted to share.

A little linguistic theory

There are two linguistic theories that are relevant here. First, is that language is used for a purpose and the type of language used is a reflection of that purpose. Which is why an academic paper will use entirely different type of language than a text conversation. The medium is different. The formality is different. The relationship between participants is different. All of that is reflected in the language.

The second theory of interest is that linguistic features don’t occur in isolation and certain features are more likely to be found together. Some of these co-occuring features are obvious. Texts with more verbs are also likely to have more adverbs. Some are less obvious.

We can combine these two theories into a practice that helps us analyze writing by looking for clusters of related features and determining what purposes they serve. For instance, writing with a lot of noun-related features (nouns, articles, and pronouns, for example) tends to be more information-dense. A textbook will use more of these features than a casual conversation. In this project, I was looking at the different clusters of linguistic features appearing in successful versus unsuccessful query letters.

Some (very simplified stats)

There are two stats tests of relevance here. One is factor analysis, which is basically the statistical method for finding which linguistic features work together. This is a very involved multi-variate statistical test. You don’t need to understand it. I promise it exists.

The other statistical test is a logistic regression. This is another multivariate statistical test that basically helps identify what features predict a binary outcome, in this case whether or not a query letter is successful.

In short, I used factor analysis to identify features that co-occur in the query letters and a logistic regression to determine which features correlated with whether a query letter is successful or not.

My methodology, not that you probably care

I (ethically and with permission) gathered over 250 query letters from the agency I used to intern at. I categorized them as successful if the agency requested the manuscript and unsuccessful if no request was made (successful query letters didn’t necessarily receive offers of representation, just a request. I used a program developed by Dr. Doug Biber to tag the query letters for a variety of different linguistic features (things like parts of speech).

Here’s what I found

There were two sets of features that predicted whether a query letter is successful or not. And, as someone who’s worked with a lot of query letters, they made quite a bit of sense once I thought about them.

The first set of features that distinguished the two sets of letters included infinitives, prediction modals, conditional subordination, necessity modals, and possibility modals. These features all appeared more frequently in successful query letters than unsuccessful query letters. Traditionally, these are features are associated with overt persuasion. You should do this or if you do this, you need to do this. However, that doesn’t seem to be how they were used in query letters. Instead, they’re being used to establish character motivations and stakes in successful query letters. Some examples with the relevant features bolded:

To save the prince, Vlad must find a cure for his vampirism or risk losing the love of his life forever.

Most people want to win the tournament for cash prize, but Vlad sees it as a way to finally get the respect he deserves.

If she doesn’t get to her girlfriend before the criminals do, she stands to lose more than just the love of her life.

As you can see, these features correlate strongly with establishing character motivations and stakes. These are important in the query letter, so it makes sense that we see these features more in successful query letters.

The second set of features that occurs more in successful letters includes present tense verbs, adverbs, activity verbs, coordinating conjunctions, all conjunctions, third person pronouns, and stance adverbials. On other hand, features that occurred more in unsuccessful letters included Adjectives, nominalizations, prepositions, common nouns, and all nouns.

That’s a lot of jargon, but basically, successful query letters were verbier (not longer, having more verbs and features related to verbs). Unsuccessful letters had more nouns. What does this mean? It means that successful query letters focused more on plot and unsuccessful query letters spent more space on talking about characters and setting.

In conclusion

Looks like the science pretty much says what all the experts say. Focus your query letter on plot and make sure that when you do focus on characters you’re focusing on motivations and stakes, rather than just listing attributes.

The post What Makes a Successful Query (according to science) appeared first on Captain (Query) Hook Editorial.


Viewing all articles
Browse latest Browse all 86

Latest Images

Trending Articles





Latest Images