May we have Santa's ID, please?

It is the season. And as my son has now come to an age where he enjoys being read stories which go a bit further than simply “apples are round”, we have started reading Christmas stories together. With books we got from our local library, of course. Specifically, ‘De grote kleine kerstman’ (a translation from the book series “Der kleine Weihnachtsmann” by Anu Stohner.

In this story, there’s a village full of Santa Clauses, and the littlest one (our protagonist) is the one who delivers presents to the animals. After some miscommunications and a plot-twist, it turns out our little hero is next in line to become the main Santa Claus! This, to me, raised the question – ‘How do we know who Santa Claus is, exactly?’.

Now, I don’t want to learn about the magical character associated in modern Anglo-Saxon Christmas traditions, derived from a mix of various local, pagan Germanic Christmas figures as well as Saint Nicholas (with his own set of different celebrations across Europe). No, I mean if we refer to Santa Claus, to whom are we referring?

When we want to refer to a specific person, we simply use their names. But what if we have a collection of similarly named persons? Like John Smith or Spartacus? We can point to that person to make things clear – but when we are not in the position to point to someone, what can we use instead?

Well, then we will have to venture into the realm of identifiers. Most of us have identifiers of all kinds. Governments prescribe us with some sort of social security number to help various (governmental) organizations identify us. When reading this post, your computer will have some sort of IP address to help identify it on the Internet (simply put). Your social media page will have a URL/URI, which is an identifier as well to help identify that page - and thus the representation, or manifestation of you within that platform’s database.

Now, I was going on about Santa Clauses, so how do I use identifiers to distinguish them? If I were to set up my own database on Father Christmases, I would give each manifestation representing them an identifier to help distinguish them. This would work in my database, but not within other databases. To do that, I would have to map every Santa Clause instance from both databases and refer to each corresponding identifier. That way, whenever the datasets need to communicate, we know we are talking about the same Santa. When a third database were to pass me by, I would have to repeat this process again, and again with a fourth. A different approach would be to use a single database as a type of common denominator – as a type of authority.

One of these common identifying sources, VIAF (Virtual International Authority File), has a representation of Santa Claus. Also of another Santa Claus, which exemplifies the problem of similar names mentioned earlier. Other commonly referenced databases also include Santa. For example, Wikidata. Notice how it refers to (via linking) the beforementioned VIAF, stating it is the same person. We can also spot him on the French Bibliothèque nationale de France (BnF), the German Deutsche National Bibliothek (DNB/GND), or WorldCat Entities.

But this isn’t quite the clear-cut chocolate cookie and cup of warm milk as it initially looks to be. Yes, we can identify various manifestations of what seems to be the same thing, and yes, we can link these to say ‘hey, our thing is the same as your thing’, but if we look closer, we notice that something looks a bit off.

For example, VIAF says that an alternative name for Santa Claus is ‘Kris Kringle’. VIAF is not alone here, as Wikidata does the same. However, Wikidata also acknowledges another Christmas-figure as Kris Kringle, namely the ‘Christ Child’. Wikipedia’s link to DNB’s Weihnachtsman, does not provide (the Germanic) Kris Kringle as an alternative name. Instead, the DNB’s entry for Christkind does. And we can do the same of the alternative name ‘Saint Nicholas’ – which opens a whole other can of worms.

So, what does this mean? Are VIAF or Wikidata wrong? Is DNB wrong? The answer isn’t quite straightforward. It could very well be that some of these entries need some editing. For example, hopping through Wikidata, I found an entity which has as a Dutch label ‘Saint Nicholas’ - and that isn’t a Dutch label. But more importantly, Santa Claus is a unique anomaly which brings another aspect in how different knowledge representations (such as found within datasets) work.

Cover you children’s ears, but Santa Claus isn’t a real person. Rather, he is more of a conceptual figure. A conceptual figure which is also a culmination of various other conceptual figures. This makes it very difficult to create a singular consensus on who Santa Claus is and who he isn’t. Better said, what parts of the beforementioned other figures are considered ‘part of’ Santa Claus, and which aren’t?

For example – it is mostly accepted that Santa Claus s a reimagining of the Dutch ‘Sinterklaas’ (try to say Sinterklaas out loud when you’re not Dutch, and ‘Santa Claus’ is what you’ll most likely say instead). As well as Sinterklaas being a characterization of the Catholic Saint Nicholas, and so is Santa Claus (“…in hope that Saint Nick soon would be there”). However, although the Dutch have also embraced the concept of Santa Claus besides Sinterklaas (as ‘Kerstman’), there can only be one St. Nicholas bringing children gifts in December. In other words, a Dutch database will more likely make a clear-cut distinction between the two characters, whereas a different database would (due to their use cases) most likely see these as the same – be it as alternative labels or as linked manifestations.

And that is, albeit confusing at times, just fine. We all see the world, and Santa Claus within it, differently. It makes sense for those creating knowledge representations in databases to shape that knowledge to the views they have. What we must be cautious of, when linking or referencing to other data, is how that data is represented and how we want to link to it. After all, if we were all representing the same data in exactly the same way – there would be very little use for us to work in different datasets to begin with, let alone refer to each other’s data. Admittingly, when writing this, I’ve been cutting corners, and not double-checking to see whether the links between the various Santa Clauses use the semantically narrow ‘same as’ or semantically broader ‘similar to’ (or something of such a nature). However, experience does teach me that datasets are prone to use the former as a default, and not always consider the latter as an option. This does not mean that these databases do so, but many others do.

Now then, what should we take away from my post here? Although I didn't uncover which Santa is which, but what I have hoped to convey here in a ‘festive’ manner is that representations of things in datasets use identifiers to distinguish them from other, sometimes similar in name or shape, things. But also, that these representations, albeit of the exact same thing, are not always represented in the same way. When looked closer, differences (and in some cases contradictions) appear. These differences are not something to be worried about – in fact, they can be very useful and even necessary from particular perspectives – but it is good to be mindful of these differences when linking your data to other data.

And that’s it. Feel free to comment, react or reach out to me more directly if you want to share your thoughts on the manner.

Happy holidays everyone. Keep safe, keep warm, and let’s have a great 2023!