Tinderbox Forum

Talmudic-style composition: how to represent layers of authority?

The Talmud is often printed in such a way that multiple commentaries on a text appear around that text, but in a way that visually distinguishes them as less authoritative than the central text, viz:

My interest is not biblical, but I wanted to offer this as a visual. What I am interested in is creating a Tinderbox document composed of multiple layers, where each layer has a different degree of authority. (Layer is not meant in a literal sense, but I haven’t found a better descriptor.) The base layer of the document would be the most authoritative information, with additional layers stacked on top adding additional information that is less authoritative than the layer it rests atop. In the simplest non-trivial case of two layers, the base layer would constitute an authoritative version of the document, and the other layer would provide additional information to the base layer, without permanently obscuring it. The physical metaphor here is a stack of transparencies, where the transparencies on top can be removed to look at and make changes to the transparencies lower in the stack.

The central use that I’m driving at both with the Talmud visual and layer metaphor is that I would like to have a Tinderbox document that includes a great deal of information taken from authoritative sources and organized to my taste. I then want to add my own thoughts and amendments without losing the ability to add to and refer to the underlying document composed of authoritative text without the changes I’ve authored.

I have only scratched the surface of what Tinderbox can do, so I thought I would ask the forums: is any of this something that Tinderbox supports? Or am I better off writing tools to manipulate the underlying XML in the ways that I’m trying to accomplish?

Thanks for your advice.

(Image from https://commons.wikimedia.org/wiki/File:First_page_of_the_first_tractate_of_the_Talmud_(Daf_Beis_of_Maseches_Brachos).jpg)

1 Like

I assume that you are not looking to use containers and outlines? That structure, inherent to Tinderbox documents, could be used for the layering you mention – and very often is used that way by many people.

Since you refer to the Talmud structure and layers of authority (Mishna, Gemara, Rashi, etc.) I assume you want this all to be on the same level. Perhaps maps with adornments to represent the “layers of authority”?

The word ‘printed’ is very pertinent here. This does feel akin to the story of railway gauges and a (supposed) link to the width of roman chariots. The image above is printed but as I understand it is essentially a print facsimile of an annotated manuscript. Be it papyrus, parchment, or other contemporary media, they had size limits, so annotations on the main text—thus-named marginalia—had to be written in the margins around said text. Latter annotations fitted around earlier ones in a pragmatic use of the available physical unused page. Thus there may be first order, second, third, etc., levels of annotation. The image shown above are, at base, simply margin notes upon margin notes.

So, what’s unclear to me, in a Tinderbox context, is if are you really looking to create a digital visual facsimile of a print convention, or to update it for the digital medium? That question might appear odd at first, but I think the two take you in different paths. Worst would be to set off down the wrong path.

The facsimile path implies a map-based view with link-based selection of placement of notes. The ‘dance’ of force-directed layout may be useful. However, unless you have an enormous screen, you will soon find yourself with maps that are too large to viewed whole at a useful degree of zoom. What if an annotation relates to something a different ‘page’? Our erstwhile scribe/printer would have to have added a “see…” annotation. Digitally we can do this with a link. But where then does the original annotation sit? On the first ‘page’ where it occurs? On every page it references? As a separate instance of the annotation, or as an alias to a separate canonical reference … stored where?

A more hypertextual approach offers all sorts of potential. Indeed, part of the concept of hypertext is to break out of the mechanical limitations of the paper page, and from linear writing methods that don’t allow for branching and diversions. This isn’t a binary good/bad choice but simply a recognition that the paper page is limited in the amount that can be printed upon it.

Informational layer or visual layers? Either way link types and metadata (attributes values) are likely to be part of the solution.

It is the case that at time of writing, maps don’t offer visual layers. Before someone notes map items visibility can be toggled, this is true but from past experimentation I’m not sure this suits here. You’ll rapidly get tied up in visualisation control code when what you actually want to do is see the relationship of text and notes.

My hunch is the comparatively new hyperbolic view is what you should investigate. This view shows only linked items, so you will be making lots of thinks—though as this is deliberate work it is good as you are thinking about the attributions (links) you are creating. By centring the view on a note (core text) you can look outwards at all the notes (annotations) that point backwards to it.

In Tinderbox there’s no one ‘right’ way to do things - it’s a big toolbox rather than a set of templated solutions as in many other apps. All views are simply different visualisations of the same notes. Under the hood, a document is all just links, notes and notes are a cluster of attributes (all attributes are available in all notes even if not used).


I always learn something from your answers, even though I didn’t ask the question. Thanks Mark.

1 Like

Daniel Bomberg is an interesting figure – he bequeaths not only the spatial layering of commentary around a central text, but also the chapter and verse segmentation which we take so much for granted.