Categorizing Time

The author of the Pisan Chronicle Add MS 10027 relied heavily on a variety of time-based indicators - what we have called temporal markers - to construct the historical narrative of his city and move his story line from one event to the next. As readers make their way through the text, the variety or types of temporal markers employed by the author in turn invite a series of questions such as, “When does the author use a liturgical temporal marker instead of a secular one?” or “What are the implications of defining time in relation to a person or group instead of a calendrical date?”

In some parts of the chronicle, temporal markers of vastly different scope are piled one on top of the other to inform readers about what occurred in the story at varying granularities of time. The frequent reference within the manuscript to state officials like consul, capitano di guerra, or podestà, and the length of time they remained in office, suggests that temporal markers might also function to demarcate an extended period of time (duration) as opposed to pinpointing a specific moment, what this project has termed a timestamp. At times, temporal markers referred back to a person or event mentioned previously in the text, so that they relied upon the internal logic of the text itself, a function the project describes as self-reference.

By no means exhaustive, the system of categorization was formed in response to the questions that developed during the transcription phase and is predicated on a hierarchy of subject classifications. While there are many criteria by which temporal markers might be organized, these three in particular-- type, function, and granularity--best reflect the temporal organization employed in the text by the author; that is, his way of reckoning time. By tagging each of these marker found in the text using FromThePage, a master list with the tagged subjects was compiled as a dataset and then used to search for patterns of temporal-marker use over the whole chronicle.

Initial Observations

A few simple visualizations reveal some basic information about how the author used temporal markers in his narrative.

First, the total number of temporal markers fluctuates greatly from page to page; often, a page with a large number of temporal markers (such as ff. 8r, 10r, 34r, 35v, 37r) are flanked by pages with few or none.

Also, each category has one class that comprises the vast majority of the temporal markers: timestamp is the most common type of marker for function, year for granularity, and calendrical marker for type, as is made clear in the color-coded charts below. The analysis thus focused on concentrations of classes other than these powerhouse players: such as the appearance of liturgical markers on f. 15r and ff. 37r-39r, and the concentration of self-reference in ff. 7v-10v.

As useful as these preliminary visualizations are for orienting the trajectory of our analysis, they nevertheless have their shortcomings. Foremost among them is that they only address distribution by page, not by the narrative organization of the text. To get the most accurate read of the temporal layout of the manuscript, we must respect the narrative units of the text as we conduct our analysis. Thus, it was important for us to pivot away from thinking of our manuscript in terms of pages and toward imagining it terms of episodes.
When adjusted to reflect the lengths of each episode and the number of temporal markers within it, new trends emerge that allow us to narrow our focus even more and select certain portions of the text for closer study. The following page will explain in more depth the way we defined and identified these episodes, and the usefulness of an episode-conscious approach to the text.