SearchResearch is about learning...
 |
| St. Peterskirche, Zürich |
...how to answer those questions that come up and make you wonder. As mentioned, this happens a LOT while traveling, so this week's quest is to learn how to answer those kinds of popup questions.
1. In the picture of St. Peterskirche (above) there's something about the clock tower that struck my eye and made me say, "That's funny..." Does it strike your eye too? Can you find an explanation for it? (No, it's not leaning.)
The thing that caught my eye here is the Roman numerals on the clock face. It's I - II - III ... and then ... IIII.
Surprise! It's NOT IV, as I expected.
As I travel through the world, I'm always looking for things that are not what I expect. In this case, I fully expected the clock face to be I - II - III - IV
As I wandered around central Europe, I found a LOT of clock faces with the IIII numeral.
So what's up with using IIII rather than IV?
I started with the simplest query:
[ IIII on a clock face ]
And found a number of fascinating points. While I saw the IIII as assumed it was a central European thing, this query showed me that it's used on most clock/watch faces that use Roman numerals.
And it's in common use in the US as well:
 |
| Grand Central Clock |
But there are notable exceptions, such as Big Ben in London, which uses the IV numeral.
If you just look at Images with the query:
[ clock face roman numerals ]
you'll see that the sample is pretty evenly split between IIII and IV.
But when I saw the IIII numeral in central Europe, because I was in an unfamilar place, I naturally assumed it was a Swiss or central European quirk.
This is a kind of attribute substitution inference. That is, because I was in an unusual place and saw an unusual thing, I inferred that the thing was linked to the place. That is, the use of IIII was a property of my location, and not a sometimes-followed convention around Roman numerals on clocks!
I bring this up because it happens a lot, especially when searching while traveling. The temptation is to do a query like:
* [ Swiss Roman numeral clock face IIII ]
But notice that I've baked-into the query the assumption that this is Swiss. But beware: If I ask an LLM a similar question:
[Why do Swiss clock towers use IIII rather than IV on their faces]
... you'll get a similar kind of biased answer. Both Gemini and ChatGPT give a version of an answer that doesn't undo the bias. ChapGPT: "Swiss clock towers use "IIII" instead of "IV" mostly for aesthetic balance and historical tradition."
A better AI answer would be to say something like "The use of IIII vs. IV is seen around the world. It is not a convention that is particular to Switzerland or central Europe." True intelligence anticipates the errors that might come up in an answer--that's what I'd do if answering your question.
But that's not how LLMs work. If you bake in the bias, you'll get the bias back in the answers.
However, the LLMs are good at coming up with explanations about WHY this might be. The summary of the LLM's output is basically:
1. Aesthetics and Visual Balance: The "IIII" provides a better visual balance on the clock face, especially when contrasted with the "VIII" (8) on the opposite side of the dial. The "IIII" also helps divide the clock face into three visually distinct sections (I-IIII, V-VIII, IX-XII), each using a similar visual pattern.
2. Historical Usage: While "IV" became common later, "IIII" was frequently used for "4" in ancient Roman times, especially on sundials. The subtractive notation (like IV for 4, or IX for 9) became more standard after the fall of the Roman Empire, but the older additive style persisted in some applications.
3. Ease of Manufacturing: For early clockmakers who were often casting or cutting out metal numerals, using "IIII" might have simplified the process. Using “IIII” for 4 and “VIIII” for 9 (that is, all additive numerals) would require the use of only three molds. The first “IIII” mold for casting numerals 1 through 4; the second “VIIII” mold to cast 5 through 9; and a third “XII” mold to cast 10 through 12. Adding “IV” and “IX” would have required additional molds and thus would have made for a more inefficient process.
It's difficult to find good authoritative citations for any of these arguments. There are LOTS of web pages with speculations, but the most reasonable one I've read is the "Ease of Manufacturing" speculation above. I can easily imagine clockmakers trying to simplify the process.
Plausible, but there's no clear authority for this story (or any others).
So we'll simply agree to keep looking for a deeper explanation, but note in passing that this is a worldwide phenomenon.
2. All around Zürich I kept running across places that had this logo (below). They always seem to be centered around a coffee shop / diner of some kind, but they seem to be much more than just another coffee shop. What are these places? Why would one go there?
A simple search for this text (Zürcher Gemeinschaftszentren, aka GZ) takes you to several pages in the city of Zürich website. The GZ are described as "... a network of 17 community centers in Zurich, Switzerland. They serve as vibrant hubs for local residents, offering a wide range of activities, workshops, cultural events, and spaces for community interaction. These centers are co-funded by the city and are open to the public, providing a welcoming environment for individuals and families."
You can get a sense of the activities at the GZs by looking at their home page: https://gz-zh.ch/ Read around in these pages to get a good sense for the kinds of things they do. (Note that you can set your Chrome browser to automatically translate to your preferred language: Look for a section related to "Languages" or "Translation" within your browser's settings or preferences.)
If you live near one, you'll find a coffee shop that's a place for you and your neighbors to hang out, a lunch place (often with free lunches!), and usually an activity space that's very busy. (The one near me had regular musical performances by good bands!)
You might want to go to meet up with people in the neighborhood or to take advantage of the free activities. It's well worth your time to stop by.
(In my case, I saw this logo outside of a couple of different buildings and had to look up what they were. A fascinating cultural meeting ground for people who live nearby. Check out their YouTube channel to see what kinds of things go on at the GZs.)
3. I was lucky enough to see one of these wee beasties flying over a clump of flowers and sipping nectar, looking for all the world like a hummingbird. But it's NOT a hummingbird! What is it? (I wish this image was mine, but none of my shots turned out. This is from Wikipedia.)
 |
| A mysterious flying critter. What is it? P/C Wikimedia |
A simple Search-By-Image identifies this as a Hummingbird Hawk-moth (Macroglossum stellatarum). Wikipedia tells us that this is "...species of hawk moth found across temperate regions of Eurasia. The species is named for its similarity to hummingbirds, as they feed on the nectar of tube-shaped flowers using their long proboscis while hovering in the air; this resemblance is an example of convergent evolution."
4. You know the quote above that's reputedly by Asimov? ("The most exciting phrase to hear in science, the one that heralds new discoveries, is not “Eureka!” (I found it!) but “That’s funny…”) I have my doubts. Can you figure out where it actually comes from?
In conclusion, the ascription to Isaac Asimov remains uncertain. He received credit in 1987 while he was still alive. The attribution appeared in the “fortune” program of the UNIX operating system. The quotation has not yet been found in the writings, interviews, or speeches of Asimov. An interesting precursor occurred in an article by Gordon Rattray Taylor in 1965.
