[mumblemumble]

> The underlying evidence in this case is clearly fragmentary and unclear.

At least based on what's being reported in the Atlantic article, it seems pretty clear to me: The most distinctive features of the word are inferred, and they're inferred well out past a selvedge. The presumed style of calligraphy appeared 500 years after the textile was made. And people who are familiar with the script are saying that even the parts of the pattern that we have aren't compatible with the idea that it's meant to spell Allah.

I don't have the expertise to evaluate any of those lines of evidence, except perhaps the selvedge question, myself, but, if we take, for the sake of argument, that they're accurate, that is multiple very different lines of evidence that all conflict with the hypothesis. I wouldn't be surprised if that's as clear a disconfirmation as you can typically get in archaeology.

That said, it doesn't render the basic idea risible; we know there was plenty of trade and cultural exchange, and we have plenty of other examples of artifacts with Arabic writing being found at Viking sites. It's just that it seems much more likely that this particular artifact merely exhibits an abstract motif.

[sfjailbird]

Not sure which sources you are referring, but the lettering looks like the earliest Islamic calligraphy (Kufic script, which is beautiful) [1].

This is the word Allah in Kufic:

https://freeislamiccalligraphy.com/?portfolio=allah-square-k...

The line above is ornamental. Pretty close to the clothes design.

[1] https://en.m.wikipedia.org/wiki/Kufic

[mumblemumble]

Specifically the square version of Kufic, which is apparently a more recent variant.

First, the style of Arabic that Larsson says she has identified—square Kufic—is not known to have been used in the 10th century; it only became common about 500 years later.

I imagine it's just like how, with the Latin script, anachronistic letterforms can be used as a line of evidence in identifying less-skillful forgeries.