You might be underestimating how many Jewish academics are secretly cheering with schadenfreude.
Thank you for reading.

You can create an account and read it in full for free

I enjoyed reading it, and learned a lot!

With Lara Yang, Sarayu Anshuman and Sameer Srivastava.

Do our findings extend beyond organizations? We speculate they do. The effects of global reach on identification are highest for those in very clustered networks. Identification with an "imagined community" requires a combination of local cohesion and global integration.

The problem with word embeddings is they assume all text was generated by the same model. Mittens allows us to fit models at the person-time level.
Such an approach is very useful for understanding interpretative heterogeneity across people or time!

www.annualreviews.org/content/jour...

We look at two network measures (inferred from email exchange): local clustering (how tight knit one's ego network is) and a new measure, global reach (how distributed one's ego network is across multiple communities).
Both predict identification, even within person over time.

To measure identification, we use Mittens, an extension of GloVe, to fit person embeddings (from email) across 3 orgs. We measure the similarity between "I" and "we" as each person's identification strength at a given time unit.

github.com/roamanalytic...

More fun applications in the article. You can find a pre-print here:

osf.io/preprints/so...

\end

What about schemas, you ask? Somewhat heretically (at least for cultural sociologists), we find this too expansive an analytical concept. We hope that thinking about categorization and semantic association, sociologists can be more precise when they talk about schemas. >>

Thinking about interpretation, and especially interpretative heterogeneity, is important, because the links between people’s beliefs and their actions are often mediated by interpretation. You and I might both believe in diversity, but disagree on what this ideal means. >>

Framing, for example, affects recipients’ categorization and semantic association, both through verbal (by labeling and analogizing) and non-verbal (relationally and through co-occurrence) channels. >>

To interpret something is to associate it with a cognitive concept (categorization), invoking a set of associated concepts (semantic association). This enables us to conceptualize shared interpretation, and the process by which interpretations are coordinated across people. >>

Co-authored with the wonderful Abraham Oshotse and Yael Berda

Thank you. Given that I adopt technological trends past their peak, I worry if this means bad things for bluesky....