Latour suggests using four notebooks when carrying out an ANT (actor-network theory) study. They are listed in his chapter on the Fifth source of uncertainty:writing down risky accounts. The principle is to keep track of every move because "everything is data". This separation of different types of data (related to different kinds of research activity) will also aid retrieval and shuffling data.

The Four Notebooks
==================

1) To log the enquiry.
Appointments, reactions, meetings, times, dates, locations, technologies used.

2) For gathering information.
To be arranged simultaneously chronologically and by category. He suggests the use of digital file managing software for this notebook, to allow a shuffling between different arrangements of data.

3) A sketchbook- for first drafts of writing
Written notes and sketches are to be made throughout, and at every stage of, the study as the research activity occurs. This will provide the material for proper reporting, ensuring that the vocabulary and substance of the record remains fresh and is not generalised into a series of cliches by memory. A place for uninhibited thought - this is where you work through unhelpful attachments to metaphors and disentangle in your mind the metalanguages of the actors and analysts.

4) To register the effect of the written account on the actors whose world has been either "unified or deployed" - to check "how the account plays its role in assembling the social"

Notes on Reassembling the Social (Latour 2005)