search

SunoikisisDC / SunoikisisDC-2020-2021

Sunoikisis Digital Classics 2020–2021 syllabuses
19 stars 2 forks source link

Discussion of session 4 readings: Palladino 2016 + Simon 2019 #25

Open gabrielbodard opened 3 years ago

gabrielbodard commented 3 years ago

Please think about provenance (who are the authors, what is their relationship to the work they're discussing, where is it published), and how these discussions fit into the questions about gazetteers and representation of place and space that we've been thinking about the last couple weeks.

nicolealexandra33 commented 3 years ago

Rainer Simon, Valeria Vitale, et al. 2019. “Revisiting Linking Early Geospatial Documents with Recogito.” e-Perimetron 14.3, 150-163. ‘Recogito is an open source tool for semantic annotation’ which allows for a great number of applications and techniques for collaboration and is maintained by the Pelagios Network, a ‘non-profit association advocating and enabling the production and use of linked open geo-data in the humanities and cultural heritage’. Recogito was first introduced in 2015 where it was only used internally on a project called Pelagios, but received a great deal of attention subsequently. This encouraged the development of the newer version of Recogito which included a friendlier user interface suited for both non-technical and experienced users; and greater support for general annotation (namely marking up of people, places, things, etc.) and TEI texts. Besides annotations, another useful function for which it allows is the digitisation and markup of maps. Once they are digitised, it also allows for the map data to be converted fairly easily, in order to be exported via different software. Much like other hosting programs, there is also a tab for downloads as well as documents and controls for collaboration levels. However, the authors acknowledge that Recogito can be used in a number of unique, new ways, and as such, the platform is constantly evolving to include new tools. Some of the best known projects which have been made possible with Recogito include ‘The Lazarus Project,’ ‘Digging into Early Colonial Mexico’, ‘Kima’, and ‘Pelagios al Sur’. It has also proven to be a very useful teaching tool, especially with undergraduates who are learning new concepts and being introduced to ancient sources, as Recogito can display such information in a more engaging way.

PAZHames commented 3 years ago

@nicolealexandra33 thanks for the summary! You're so right, it would be really interesting to hear about Recogito's use in more archaeology-based projects. As they only mention text-based examples (including source maps), I assumed this was the primary/sole purpose of the project. In what way were you thinking it might be most useful for long-term archaeological projects?

ChantalvanEgdom commented 3 years ago

Chiara Palladino. 2016. "New Approaches to Ancient Spatial Models: Digital Humanities and Classical Geography." In Digital Approaches and the Ancient World. Edd. G. Bodard, Y. Broux & S. Tarte. BICS 59.2, 56-70.

Defining ancient geography

Ancient geography had a periplographic nature, being primarily conceived in the shape of a practical narrative, shaped by the establishment of travel-routes. During Alexandrian and Roman times, there was a shift towards a more explorative and structured method of organization of knowledge. Philosophers like Thales and Anaximander were the ones to provide another point of view, namely that of space as a geometrical entity. This resulted in a double track of geographical perspectives, with both empirical and abstract perspectives existing alongside each other, which did not come together to form one method. Only recently has it been acknowledged that both of these systems are worthy systems of shared knowledge, constantly developing with regards to sources, modifications, reprisals and corrections.

Digital humanities and ancient geography: what has been done, what could be done

Graeco-Roman geography-studies have become one of the most active in recent years. There is a substantial lack of consensus on even the most fundamental topics, which led to the willingness to interact with new methodologies, also due to the fact that spatial research fits well into large-scale approaches and benefits from innovative visualization and mapping systems.

• Creating GIS systems in 2000, the first comprehensive coverage of Ancient Mediterranean geography, using the Barrington Atlas. More improvements were made at the Ancient World Mapping Center, which led to first web app for generating interactive maps being made. Improvements set the stage for later important geographical gazetteers, of which the Digital Atlas of the Roman Empire and Pleiades (which produced a new way to conceive space through Linked Open Data). Notions of space were redefined as conceptual entity.

• Further research focused on spatial practices. The Stanford ORBIS project researched the conditions of travel in the Roman world.

• Tabula Peutingeriana redefined importance of cartographic perspectives. GIS-systems integrated with primary evidence maps.

• The Hestia-project can be considered to be the most important in textual research from a methodological stance, challenging modern special perspectives: space understood as lived experience, rather than abstract topography.

Focus on the problem

Three key-concepts in Graeco-Roman geographical narrative: distance, orientation, spatial semantics; three datasets, functioning to clarify structural specificities.

• Distances: concept of distance depended on concrete conditions of travel and were structurally unstable, computed in units of time. When used in narrative, an increasing preoccupation for standardization began, with Herodotus being the first to provide consistent measurements. However there was some arbitrariness and approximation. Units of space were unstable, conversion between space units were never uniformly adopted, and time was still used to calculate space as well. The Roman Empire brought organic and stable communication networks and structural systemization, introducing “the mile”/1000 paces (never the exclusive practice). Spatial measures not very prevalent in Latin geographical texts, but widely used in division of land, which demonstrates strong connections between geographical information being derived form external contexts and geographical information preserved in documents.

• Orientation: ancient sources show orientation on two main references: o Formalized points (compass points, rose of winds)/conventional values: static locations, sometimes elements of astronomical lexicon, especially in poetic/aesthetic texts, in which case it often expressed motion. o Unfixed points (buildings, natural features)/relative values: overlapping issues of spatial semantics, but two layers are to be distinguished: purely linguistical aspects and issues of definition. Important is the connectivity between spatial entities in the description of space. The problem with linguistic encoding is that the understanding of even simple indications of position are only apparently straightforward. (as noted by Alexander Podossinov, simple prepositions like ὑπέρ, ἀπό, sub, and supra cannot be associated to objective orientation points).

• Semantics of space: conceptual and directional narrative serve different purposes and have different expressive structures, hence separation from linguistic expressions of orientation as mentioned above. Conceptual space has cognitive and social meanings, rather than a strictly descriptive one. Directional expressions function more empirically. o There are two types of spatial semantics in the article: landmarks and concepts/categories. Landmarks help articulate human perceptions of space, often culture specific. Categories are cultural artefacts (definitions of space in terms of shared knowledge). Concepts are ideas that characterize objects. Spatial categorizations have deep cultural implications. Establishing a methodology in Digital Humanities There is a crucial need for formalization and standards. Now, there is no predefined interpretative model: a bottom-up approach to create standard is needed. The article suggests to establish analytical and exploratory methodologies, using two data-sets: language-centered and concept-centered.

• Linguistic analysis of spatial orientation: language as a structure, which shapes description of space. The reconstruction of linguistic structures can be addressed through linguistic annotation, following standards provided by the AGLDT project, a linguistic treebank. A corpus driven, bottom-up collective of data, which can help understand the evidence through associations of expressions of spatial orientation with linguistic constructs.

• Annotating distances and directions: ancient geography relies on cognitive processes, where navigation is defined through special relatedness. The absence of predefined vocabulary demonstrates the need for annotation workflow of marking structural linguistic patterns. Place-relatedness can be expressed and encoded through triples: subject, predicate and object. RDF offers a generic abstract data model for doing this, using predefined vocabularies, or ontologies. The GeoLat project offers opportunities to explore ontologies of special mentality in Latin texts, focusing on geographical content of single texts, displayed on GIS maps, but will probably provide a semantic aid to the systematization of geographical concepts.

However, predefined vocabularies can be problematic, as the standard can exclude certain points. On top of this, in space-relatedness the standard is absent. This calls for a suitable digital method for annotation of meaningful linguistic structures, without superimposing a predefined semantic model. The mARkdown system is such a model, providing a highly articulated system for the annotation of meaningful patterns. It has an easy and customizable range of tagging options, with potentially infinite granularity. It demands minimum modification of the textual data and challenges to recognize underlying linguistic structures. Three basic types of patterns can be distinguished: distance, formalized direction and unformalized direction, providing an initial conceptual basis.

Annotating spatial semantics

Association between concepts/categories and place names is fundamental. In this process, Pelagios, a community-driven initiative, connected to Linked Open Data, is important. It recognizes common references rather than common vocabulary, through places as the basis of an interconnected world. Recogito was developed in order to support the exploration of new research techniques. It lacks a semantic model and is particularly useful for free-scheme annotation and bottom-up strategies, based on named entities and related features. The goal is to move towards an increasingly specialized vocabulary, or semantic annotation. Landmarks can be measured through visual points, using modelling of already annotated data in a topological representation, and through data-mining, using measurements of connectivity and frequency based on the occurrence of single nodes and their relations from annotated texts.

Conclusion

This method is mostly designed to be used on a specific corpus, especially the Geographi Graeci Minores, especially in the first stages of data gathering. The results will be able to be used in larger contexts and the method is hopefully able to broaden horizons of research.

A very intriguing article! I hadn’t previously considered the differences between the ancient way of experiencing space and the “modern” way, and I think the article provides good ways of exploring the problems that have previously held back such research.

nicolealexandra33 commented 3 years ago

@PAZHames It's a good question regarding how it can be more accommodating to archaeology projects. I do know that they try to distinguish ancient and modern settlements and places, but I do think having a bit more control over where to put a place marker is one thing that would be helpful for archaeological projects. For example, when I was attempting to annotate a bit of Strabo's Geography, I could get in proximity to where I would want to put a marker, but not exactly where an ancient lesser known settlement/geographical feature is

FabioDFernandes commented 3 years ago

@PAZHames It's a good question regarding how it can be more accommodating to archaeology projects. I do know that they try to distinguish ancient and modern settlements and places, but I do think having a bit more control over where to put a place marker is one thing that would be helpful for archaeological projects. For example, when I was attempting to annotate a bit of Strabo's Geography, I could get in proximity to where I would want to put a marker, but not exactly where an ancient lesser known settlement/geographical feature is

I would imagine that it becomes more limiting to not be able to annotate more precisely on the maps. I think in the context of archaeology (which is not where my forte lies), a tool like this, and the ability to annotate more precisely, would be useful in documenting certain trends of distribution of certain things (over long geographical distances), or in an urban archaeological context (i.e. Pompeii, Ephesus), perhaps tracing on an interactive map how the distribution of material finds might tell us about each part of the city, such as the function or any social interpretations. Though, in my still limited understanding of Recogito, perhaps that would require a more powerful system? I do sense that its purpose is to trace geographic data over a longer span of distance than with such preciseness, which is why texts and maps receive the main attention.

FabioDFernandes commented 3 years ago

@ChantalvanEgdom Thank you Chantal for your very detailed summary! It certainly is interesting how the ancients perceived space (and thus travel, geography and the world) differently to us, and it was good noting how there was a general empirical geographical perspective ('you follow the south wind, look for the lighthouse,' etc.) as well as an abstract, which I suppose was likely more the reserve of philosophers like you mentioned. The linguistic aspect seems to me as though it would be a particular challenge, if I've understood it correctly - how exactly linguistic representations of space and travel from the time, as vague as many of them could be (beyond famed travel logs), can be portrayed in a digital context. Regarding the annotation of distances, I imagine that further development here would obviously be useful in the study of long-distance communications, travel, migration, trade, and have wider implications on matters of the economy and society of large open areas such as the Roman Empire. I have come across the ORBIS map before, and it is certainly a very ambitious project and pertinent to these things.

K-Doering commented 3 years ago

It certainly is interesting how the ancients perceived space (and thus travel, geography and the world) differently to us, and it was good noting how there was a general empirical geographical perspective ('you follow the south wind, look for the lighthouse,' etc.) as well as an abstract, which I suppose was likely more the reserve of philosophers like you mentioned. The linguistic aspect seems to me as though it would be a particular challenge, if I've understood it correctly - how exactly linguistic representations of space and travel from the time, as vague as many of them could be (beyond famed travel logs), can be portrayed in a digital context.

Thank you @ChantalvanEgdom for your great summary! Thank you also, @FabioDFernandes, for your insights!

I am not sure about the UK but in the US at least the prevalence of geographic semantics is still quite strong by region. Where I am from, in the Midwest, we measure everything by time and proximity to town/neighborhood (a place is always measured by it's distance in minutes driving and distance/proximity to town landmarks).

I agree with Fabio that measuring these regional, community-based linguistic differences and statistical inequalities would be a challenge. I think the annotation process is moving us closer to understanding the spatial nuances of the ancients but it is difficult enough to measure the differences of communities that are living today. My final thought would be the trouble of geographically annotating works like the Aeneid as space and time in the epic are still being debated as we type, such 'unfixed' spatial issues are mentioned in the 'Orientation' section of the article. It is difficult to see in what ways these challenges can be addressed.

PAZHames commented 3 years ago

@PAZHames It's a good question regarding how it can be more accommodating to archaeology projects. I do know that they try to distinguish ancient and modern settlements and places, but I do think having a bit more control over where to put a place marker is one thing that would be helpful for archaeological projects. For example, when I was attempting to annotate a bit of Strabo's Geography, I could get in proximity to where I would want to put a marker, but not exactly where an ancient lesser known settlement/geographical feature is

I would imagine that it becomes more limiting to not be able to annotate more precisely on the maps. I think in the context of archaeology (which is not where my forte lies), a tool like this, and the ability to annotate more precisely, would be useful in documenting certain trends of distribution of certain things (over long geographical distances), or in an urban archaeological context (i.e. Pompeii, Ephesus), perhaps tracing on an interactive map how the distribution of material finds might tell us about each part of the city, such as the function or any social interpretations. Though, in my still limited understanding of Recogito, perhaps that would require a more powerful system? I do sense that its purpose is to trace geographic data over a longer span of distance than with such preciseness, which is why texts and maps receive the main attention.

@nicolealexandra33 I can imagine that would be really helpful - a really interesting point! When annotating for Strabo, did you feel like that lack of specificity held you back at all? Or was it ok as primary targets were bigger?

@FabioDFernandes I think you're right - it does seem to make most sense focusing on texts and maps for that reason. Maybe this could be a new off-shoot for Recogito, or a separate team to build on their excellent work?

RebeccaKimberlin commented 3 years ago

Thank you both for your summaries! I would also add to Despina's point on geographic semantics, they are certainly prevalent in the UK too - I immediately thought of online walking tours / blog posts where the instructions often say 'walk until you reach x pub' or 'turn at x landmark' which similarly have the effect of mapping a place not through precise geography but physical, tangible reference points (that wouldn't necessarily be clear to someone looking at the area in, say, hundreds of years time!)

FabioDFernandes commented 3 years ago

@FabioDFernandes I think you're right - it does seem to make most sense focusing on texts and maps for that reason. Maybe this could be a new off-shoot for Recogito, or a separate team to build on their excellent work?

Yes, it would certainly be an area for further development in that aspect, and it would be a useful tool in the archaeological context. I wonder if anything like this has already been designed.

K-Doering commented 3 years ago

@PAZHames It's a good question regarding how it can be more accommodating to archaeology projects. I do know that they try to distinguish ancient and modern settlements and places, but I do think having a bit more control over where to put a place marker is one thing that would be helpful for archaeological projects. For example, when I was attempting to annotate a bit of Strabo's Geography, I could get in proximity to where I would want to put a marker, but not exactly where an ancient lesser known settlement/geographical feature is

I would imagine that it becomes more limiting to not be able to annotate more precisely on the maps. I think in the context of archaeology (which is not where my forte lies), a tool like this, and the ability to annotate more precisely, would be useful in documenting certain trends of distribution of certain things (over long geographical distances), or in an urban archaeological context (i.e. Pompeii, Ephesus), perhaps tracing on an interactive map how the distribution of material finds might tell us about each part of the city, such as the function or any social interpretations. Though, in my still limited understanding of Recogito, perhaps that would require a more powerful system? I do sense that its purpose is to trace geographic data over a longer span of distance than with such preciseness, which is why texts and maps receive the main attention.

@nicolealexandra33 I can imagine that would be really helpful - a really interesting point! When annotating for Strabo, did you feel like that lack of specificity held you back at all? Or was it ok as primary targets were bigger?

@FabioDFernandes I think you're right - it does seem to make most sense focusing on texts and maps for that reason. Maybe this could be a new off-shoot for Recogito, or a separate team to build on their excellent work?

Just adding onto your points @PAZHames, I think an integration of a Layers of London type capability of being able to overlay era-specific or skeletal maps could be extremely helpful in pinpointing more specific archeological points. Such as an outline of ancient Pompeii that can be viewed in tandem with a modern map. We will not always be so lucky with well-preserved sites like Pompeii and Herculaneum.

Kiamanx commented 3 years ago

I would imagine that it becomes more limiting to not be able to annotate more precisely on the maps. I think in the context of archaeology (which is not where my forte lies), a tool like this, and the ability to annotate more precisely, would be useful in documenting certain trends of distribution of certain things (over long geographical distances), or in an urban archaeological context (i.e. Pompeii, Ephesus), perhaps tracing on an interactive map how the distribution of material finds might tell us about each part of the city, such as the function or any social interpretations. Though, in my still limited understanding of Recogito, perhaps that would require a more powerful system? I do sense that its purpose is to trace geographic data over a longer span of distance than with such preciseness, which is why texts and maps receive the main attention.

I feel like we are lucky to have such an accessible annotation tool that translates to mapping so well, but I agree with Fabio that a much more powerful tool is probably required for more thorough work. I heard that Recogito has ceased development, and while the site still may have a lot of problems and bugs - I feel like it has set an interesting precedent for similar programs in the future.

alexandrabushman commented 3 years ago

Thank you Nicole and Chantal for the summaries. @nicolealexandra33 I can definitely see how data from old maps could be lost and hard to be distinguished digitally. I was surprised at how easy it was to annotate on Recogito for the most part. There were some places and names that annotation was difficult but that could just be some bugs that are still on the program. Overall I think the software is very promising for helping students and people like me who are starting out in digital humanities with limited experience. @ChantalvanEgdom You bring up a very interesting point, I also never considered how our modern conceptions of space would be different in ancient times. As @K-Doering said, in the US it's very common to measure space through time (i.e. Phoenix is about 20 minutes away) rather than units of distance. After coming to the UK, I still have to think about how much closer '20 minutes away' is in London, since the ways that people travel throughout rural areas are obviously much different in urban ones. When understanding the linguistics of describing distance and location of ancient times, these are essential things to keep in mind.

despinaborcea commented 3 years ago

Thank you both for your summaries! I would also add to Despina's point on geographic semantics, they are certainly prevalent in the UK too - I immediately thought of online walking tours / blog posts where the instructions often say 'walk until you reach x pub' or 'turn at x landmark' which similarly have the effect of mapping a place not through precise geography but physical, tangible reference points (that wouldn't necessarily be clear to someone looking at the area in, say, hundreds of years time!)

I agree with you, Rebecca! Slightly off topic, sometimes it can happen that these instructions are retained hundreds of years with the actual landmark having disappeared, so it could be useful for reversed reconstruction – if we have directions for a building which no longer exists but might have been important in the past, directions can help pinpoint it on a map more precisely (of course, not 100% accurately). Similarly, I find names of streets (and toponyms in general) very interesting for their reconstruction potential. For example, the Hanging Sword Alley in the City of London has nothing to do with its name today, but apparently there used to be a 16th century fencing school there. It would perhaps be useful if this could be applied to more important historical landmarks which no longer exist.

ChantalvanEgdom commented 3 years ago

@K-Doering Very interesting to know this is still a used practice! Had no idea!

gabrielbodard commented 3 years ago

@nicolealexandra33

  • Overall, all of the authors are involved with digital humanities research and projects which accounts for their encouragement and promotion of the use of software like Recogito

More specifically, the authors are the creators of Recogito and the (then) managers of the Pelagios Project as a whole. It is self-reporting, not a critical assessment of the tool and methods.

gabrielbodard commented 3 years ago

Re measuring distance in time: how far from central London do you live?

Answering that question in kilometers makes very little sense, because depending on a lot of other factors, someone geographically closer may take a lot longer to get to University, e.g. (Is it more useful to know that I live 19km from Senate House, or that my morning commute is a 20-minute train ride?)