A homily for All Saints

Later today, I am giving a talk at the Museum of Cultural History in Oslo, which will be about remembering the dead in the Middle Ages - a topic that will also provide an opportunity to describe the cosmology of Latin Christendom, in order to explain why it was so important to remember and commemorate the departed.  As part of my presentation, I have included a picture of the Old Norwegian Homily Book, a collection of homiletic material compiled in Norway around the year 1200. The homily itself is much older, and was part of a religious heritage - or package deal - common to all the churches of Latin Christendom. Since the homily is for a feast dedicated to all the saints, the homily does not focus on any one particular saint's life, but instead provides an opportunity for compressing both biblical and post-biblical history into a short text. This homily in particular thereby provides a very valuable witness to the biblical knowledge available in Norway in the twelfth century.

AM 619 4to, f.73r

The feast of All Saints - which is on November 1 - is also a good reminder of a key aspect of the cult of saints, namely that not all saints are known to us. Consequently, new saints might appear, and old saints might become revealed to the faithful. Moreover, it is also a reminder that not all those who are venerated as saints are indeed holy. These issues are often lost in scholarship on the cult of saints, since the topic is often explored from the perspective of the political context of any given period, or with a view of the dynastic or institutional agendas of those historical actors who promoted a particular cult. However, for a phenomenon like the cult of saints to be wielded as a political tool - which indeed it was - and used to promote dynastic or institutional concerns, there has to be a genuine belief in the saints and their role in Christian cosmology. The veneration of a saint was a way of ensuring patronage from that saint, and the neglect of a saint could incur severe repercussions - there are numerous miracle stories about humans who are punished for not keeping their promise to the saint, or for not listening to the saint's instructions. Since not all saints were known, therefore, the discovery of bones that could potentially belong to a saint provided the living with a particular conundrum. If they ignored the bones, the saint might get offended. If they venerated the bones, they might end up like a community mentioned in the widely famous Life of Saint Martin by Sulpicius Severus, where the saint made a dead man speak from beyond the grave, admitting to his venerators that he was not a holy man but a criminal. It was because of this conundrum that signs were looked for, and bones were put to the test. Such mechanisms were not only a matter of ensuring elite control over a powerful religious phenomena, but also instituted to prevent misguided veneration. The feast of All Saints, therefore, is a kind of compromise, or a way to both recognise and circumvent the limits of possible knowledge. It speaks to a worldview where commemoration and recognition of holiness was important, but where recognition could be uncertain and where commemoration could be misplaced.    

The feast of All Saints begins at Vesper this afternoon, roughly around six, and for those who celebrate a liturgical office, the mystical highpoint will be at Matins, sometime in the middle of the night. As a scholar, I for my part am fascinated and intrigued by the feast of All Saints precisely because it is a feast that is difficult to imbue with institutional and dynastic concerns, and instead provides an outlet for acknowledging that at the core of a phenomenon such as the cult of saints is a very real belief in the role of saints, and an understanding of the limits of human knowledge regarding the deeds of men and women in the past. 

AM 619 4to, f.73r

 

The edge of knowledge – sixteenth-century cartography and the North

 

The sixteenth century is one of the most dynamic periods in the history of European cartography. Not only did the voyages of the era provide new knowledge that was transmitted through maps, but experimentation with projections and ways of transmitting knowledge about the world led to a wide variety of maps. However, in some cases the cartographers got things wrong. In some cases – as I hope to blog about at a later point – it was because the sources were faulty, and this is the explanation for why we find a lot of non-existent islands of some of the most cutting edge maps of the period. In other cases, knowledge got warped in transmission. It is one such case I want to write about here.      

 

Knowledge could be warped in a multitude of ways along the chain of transmission. In some cases, the knowledge could be faulty from the start, whereas in other cases something might get lost in translation, seeing as knowledge often passed through several different languages on its way from observation to the map itself. In the present case, the main reason appears to be a mixture of unfamiliarity with the territory in question, combined with unfamiliarity with the language(s) of that region.  

Septentrionalium partium nova tabula (1561)

Courtesy of the University of Tromsø's map collection

In 1561, Girolamo Ruscelli, a cartographer based in Venice, printed two maps of the North Atlantic theatre – one detailed, one rather stylised. The detailed map is known as Septentrionalium partium nova tabula, a new map of the northern parts. This map contains a lot of placenames, including settlements, rivers and islands. These names are often imprecise, suggesting how difficult it could be for those from – for example – the Netherlands or Italy to accurately render the Icelandic, Danish or Norwegian pronunciations. In the case of Iceland, this difficulty is in part suggested by the choice of using Latinised names of the island’s two dioceses, Skálholt and Hólar, rendered as Scalodin and Olensis. While the accuracy of these names can be discussed, the names themselves are correct insofar as the identification of the diocesan sees is concerned.     

 

What is more puzzling, however, is to consider the correct identification of the detailed map with some elements of the more stylised map, the so-called Schonladia Nuova, new Schonladia, issued in the same year. There are several interesting differences in the presentation of Iceland and Greenland in these two maps. The detailed map contains both names, respectively Islanda and Engronelant, the latter suggesting that information was transmitted through Dutch interlocutors. The stylised map, however, does not provide a name for Greenland – and indeed seems to suggest that Greenland is linked with the Kola peninsula – whereas Iceland is identified as Thyle, an alternative spelling of Thule. That Iceland is given the name Thule is not surprising, as these two names were used interchangeably about the island throughout the Middle Ages. Another interesting difference, however, is that the stylised map includes a stretch of frozen sea, the ‘mare congelatum’, which forms a belt of ice along the coast of Greenland.

Schonladia Nuova (1561)

Courtesy of the University of Tromsø's map collection

The most surprising difference between the two maps, however, is that the stylised map contains a strange duplication of names – a duplication of sorts. In this map, there are two place names on Iceland, namely the two dioceses. However, the names are not the same as those found in Septentrionalium partium nova tabula. Instead, the dioceses are called Skalholten and Holen. These are Danified names, and suggest that the information has been transmitted through Danish informants in Venice. A direct Danish contact would not be all that surprising, considering that several Scandinavian ecclesiastics were in exile following the Protestant reformations in Denmark-Norway and Sweden. Given the accuracy of these names, it is therefore curious to see that the Latinised and less accurate names found on the detailed map are still included in Schonladia Nuova, but not where we would expect them to be. Rather, the two Latinised names have been moved to Greenland, where they are rendered as Scalholdin and Holenses, the placements of which do not correspond to any placename in Greenland in Septentrionalium partium nova tabula. We seem to be dealing with some kind of overcorrection, where an attempt to include the vernacular names of Iceland’s two dioceses in the new map has replicated the error elsewhere. This is, of course, presuming that Schonladia Nuova was produced after the other map, which seems likely given the confusion in question.    

 

The case of the duplicated Icelandic names is a good example of how certain parts of the sixteenth-century world were, in effect, the edge of knowledge. Along this edge, some certainties existed – such as the fact that Greenland and Iceland were real places – and so did many uncertainties, such as how to accurately and correctly render the names of these places. The confusion in Schonladia Nuova is not only due to the fact that Venice is a long way away from Iceland and Greenland. It is also indicative of how Greenland was falling off the European edge of knowledge in this period. Throughout the fifteenth century the contact between the Norse settlements in Greenland and the rest of Europe dwindled and eventually stopped. Bishops continued to be appointed to Greenland’s diocesan see, Garðar, but these were bishops in name only and never set foot on the island itself. Greenland was still known, if mainly as a memory or as a landmass occasionally seen by voyagers who traversed the Greenland Sea. Consequently, instead of knowledge there was a lot of information about Greenland, and most of it appears to have been incorrect, as seen in the confusion regarding Scalholdin and Holensis. In this way, knowledge warped into information, and information was tainted by hearsay, legend, and fiction, but became common misconceptions through the medium of maps.

Schonladia Nuova (1561) (detail)

Courtesy of the University of Tromsø's map collection

Schonladia Nuova (1598) (detail)

The sea monster is a decorative embellishment in the 1598 edition of the map

Courtesy of the University of Tromsø's map collection

Reading-spots, part 5

In May of this year, I was in Santiago de Compostela for the first time. While the primary aim of the trip was to visit a dear friend, the journey was also marked by my increased fascination with the cult of Saint James the Elder, known as Santiago, one of the most famous cults of the Latin Middle Ages. The cult was responsible for a phenomenal output of architectural and textual artefacts, such as Compostela's cathedral - now mostly Baroque but with some Romanesque features retained - and the Codex Calixtinus, a manuscript containing five books that collectively are known as The Book of Saint James (Liber Sancti Jacobi). All five parts of this book have now been translated to English, and for the trip to Compostela I had brought with me translation of the first book, which is a collection of liturgical texts and sermons. The translation is by Thomas F. Coffey and Maryjane Dunn, and was published by Italica Press in 2021.  

While in Compostela, I focussed on reading the liturgy, and the book was a steadfast companion on my various excursions throughout the city. One of the most memorable reading-spots, however, was when I visited the cathedral itself, and sat down on a bench in the nave, reading through some of the texts that have been performed in this space throughout centuries. It was an awe-inspiring sensation, a feeling that heightened the experience of reading the book and also of sitting in the cathedral where the ostentatious gold and silver stand in marked contrast with the more humble witnesses to the history of the cult's lived religion, namely the various mason's marks that can still be seen on several of the pillars in the nave. 

On bad art - play, puns, and the shortcomings of artificial intelligence

But when she parted hence, she left her groome
An yron man, which did on her attend
Alwayes, to execute her stedfast doome,
And willed him with Artegall to wend,
And doe what euer thing he did intend.
His name was Talus, made of yron mould,
Immoueable, resistlesse, without end.
Who in his hand an yron flale did hould,
With which he thresht out falshood, and did truth vnfould 

- Edmund Spenser, The Faerie Queene (Book 5, Canto 1, Verse 12)

As the hype concerning generative artificial intelligence is rising and increasingly affecting various parts of society, I have found myself reflecting on the frustrating discussion about its use for the creation of images and novels. Several claims have been made regarding the possibility of using AI for writing novels, and most of these - as far as I know - fail to acknowledge that what is produced is assembled from stolen bits and pieces from existing novels, and require unethical amounts of energy to produce. Another significant point here is that images and texts assembled through generative AI are also bad - not only of a bad quality, but bad because they have no meaning to them. These products are bad art because the badness has no other element to it than badness.  

What I am writing here might at first seem both hermetic and obtuse, yet I do want to distinguish between between art that can be considered bad in terms of its quality, and art that is bad because it has no meaning. These thoughts have been offset by watching episodes of the animated TV series Batfink, a Batman parody that first aired in 1966-67, and my examples are taken from some of these episodes. My main point, however, is that no matter how one evaluates the series Batfink as art, it will always be better than anything produced through the use of generative AI. The reason for this is that the badness of Batfink - to be presented in more detail shortly - has a meaning to it, and that meaning is due to the series being created by human beings. 

 

 

Screenshot from Batfink, episode 77, The Trojan Horse Thief

Batfink was an animated series that parodied the Batman TV series that ran from 1966 to 1968, and might arguably have been intended to benefit from some of the popularity of the Batman franchise. The protagonist is an anthropomorphic bat (with features that also resemble both a cat and a mouse), who fights against various criminals together with his servant Karate, a human sidekick who follows a well-established trope of being both ethnically other and less intelligent (such as Mandrake the Magician's sidekick Lothar, for instance). The series ran for 100 episodes of around 5 minutes each, and is notable for its ludicrous plots, its pompous dialogue, its myriad of over-the-top supervillain antagonists. Moreover, the series relied heavily on recycled animation. Although this is common in a lot of animation, the Batfink series recycled its scenes with such frequency as to appear as a deliberate lampooning of the practice. When considering what animators were capable of producing in the 1960s, the Batfink series is definitely cheaply produced, and it might also be called bad due to its cheap production. 

However, even if we accept the label bad for the series, Batfink also reminds us that badness - here meaning low quality - is not without depth, and not without creativity. Because of this depth and because of this creativity, a series such as Batfink has a kind of cultural merit that can never be attained by anything produced by artificial intelligence. First of all, the dialogue, the plots, and, perhaps above all, the implausibly lucky escapes from deadly situations, are all part of a satirical tone that serves as a loadbearing beam for the series - it parodies and mocks the superhero genre, something that can only be done with intent, with deliberation. 

Secondly, the series contains a plethora of puns and references to various cultural products and to elements specific to its own period. These are deliberate details aimed at an audience expected to catch them, to understand them, and to appreciate them. While most episodes are full of these, a few examples will suffice here.  

One case in point is episode 77, The Trojan Horse Thief. Immediately, the title points to the central reference, namely the story of the wooden horse by which the city of Troy was captured, as recounted in the epic poem The Iliad, attributed to Homer. In the cartoon episode, the horse is a vehicle used by the recurring supervillain Hugo A-go-go to perform robberies. The animated horse is very different from its legendary reference point, in that it is not disguised as a sculpture, it is not given as a present, and it is not brought into a walled city in order to open its gates from the inside. Instead, the mechanical horse is used as a battering ram to break into bank vaults in broad daylight, and although it does contain a human being inside it, that human is its driver, not a contingent of soldiers. If human creativity had not deliberately parodied the legend of the Trojan horse, the appearance of a mechanical horse in an animated series would have no depth to it. Instead, there is a play on a very famous legendary episode.  

  

Screenshot from Batfink, episode 77, The Trojan Horse Thief

A second example is from episode 9, Nuts of the Round Table, where the aforementioned Hugo A-go-go uses robot knights on horseback - living horses this time - to rob banks. Once again, the title indicates the cultural reference that has inspired the plot, namely the legend of King Arthur's knights of the round table. Moreover, the pun on nuts - as in nuts and bolts - is a nod to the robotic nature of these knights. The pun and the reference are both outcomes of human creativity, and they are made deliberately and easily recognisable. 

In addition, it is worth noting that there are two further elements that might also point to topoi from the literary repository to which the stories of King Arthur belong. First of all, the horse - as seen in the screenshot below - is white, a classical image ultimately drawn from Revelation 19:11-16, where the rider on the white horse is a Christlike figure. (Importantly, this is not one of the four horsemen mentioned in Revelation 6.) The white horse is a feature of numerous medieval legends, and has remained a symbol of chivalry in modern cultural parlance. Perhaps the white colour of the animated horse is a nod to this tradition? Moreover, the automaton knight also features in various old stories, be it Talos of Greek legend (mentioned, for instance, in Argonautica), or its early modern version Talus from book 5 of Edmund Spenser's epic poem The Faerie Queene, used in this blogpost's epigraph. I do not know whether the animators and script writers of Batfink were familiar with this topos, and I do not know whether the automaton knight is a nod to this literary element. Importantly, however, because the automaton knight was created by human creativity, there is a possibility that we are dealing with a cultural reference. Had this episode been produced by artificial intelligence, however, the question would have been moot, because artificial intelligence does not have intent and can therefore not make use of literary topoi beyond what it copies and cannibalises from other cultural products. 

Screenshot from Batfink, episode 09, Nuts of the Round Table

A final example to be mentioned here is period-specific references. These are elements that only make sense because the intended audiences are familiar with them from their own contemporary frame of reference. Once such case is found in episode 11, Fatman Strikes Again (another pun which points to the very franchise that Batfink lampoons). The plot concerns a man in a fat suit who robs the safes of fat men clubs. Such clubs were once a well-known, if perhaps not widely common, feature of American social life, where fatness could be connected to status and wealth. In the twenty-first century, this reference is antiquated as such clubs are no longer a prominent part of the contemporary discourse - I do not even know if any such clubs still exist. Such period-specific references only carry meaning because there is an intent in them, they are meant for the audiences to recognise. 

Screenshot from Batfink, episode 11, Fatman Strikes Again

These examples are not many, but hopefully enough to illustrate my main point, namely that even art that is mass-produced, cheaply made, and that has rather limited entertainment value has more depth than anything produced through generative artificial intelligence, simply because the human-made art in question has intent, is deliberately bad, and contains elements that are included on purpose, and not because the elements in question just happened to be part of the trove which was plundered for parts when assembling a product through generative artificial intellgience. TV series such as Batfink might be bad art, but the badness is playful, subversive, deliberate, parodic, and there is entertainment to be had in this. This intent is what makes the series art, however bad it might be deemed to be. If the series had been created through generative artificial intelligence, however, it would not be art, but a simulacrum of art, and it would be bad because of was just that, a simulacrum without any depth, meaning or intent whatsoever.