What Does It All Mean?

Hello again. I have been very busy trying to think of editorial ideas. I even spent time looking back at my previous titles to avoid repeating myself, however most of them are silly word plays or somewhat enigmatic, so I can’t tell what I said before. This time consuming activity has put me off as usual. I have written more than 70, apparently. My attempt to get ChatGPT to summarise my previous efforts wasn’t much help. It said, “Her editorial tone is conversational yet insightful” which was nice, though hasn’t given me the information I was after. Other forms of “AI” are available. I then wasted time generating a word cloud of the titles.

We clearly see “Program much first” in the centre.

Many people dislike word clouds, and there are more accurate ways of displaying frequencies. However, the random orientation makes you analyse the graphic differently. The Christmas special version of University Challenge has given word clouds of Christmas carols as clues, and you can manage to reform the words into a well known song. Are word clouds meaningful? No, probably not. However, they are one way to represent something.

People tend to search for patterns in things, and ascribe meaning to what they see. I suspect our brains are just wired like that. Remembering a pattern is easier than recalling a full look-up table. The act of summarizing data into information either as equations, rules or patterns is a fundamental part of how we think. We can therefore delude ourselves easily, finding patterns where there are none. This propensity probably helps to fuel conspiracy theories and might perhaps be behind recent riots in the UK. The internet is littered with various articles about this, but the BBC has a reasonable write-up [BBC-1].

Riots break out from time to time, and I recall several, including watching London burn, or at least smolder slightly, from my desk at a bank in Canary Wharf for a few days. Wikipedia lists several English riots [Wikipedia-1] in date order: 1715, 1919, 1947, 1958, 1981, 1991, 2001, 2011 and 2024.

I haven’t seen all of these, of course. Now I spot a pattern. 44% end in a 1. According to Benford’s law, most numbers start with a 1 [Wikipedia-2], but ending with a 1 is a different matter. Does this mean something? Probably not.

The starting months are of note too: Oct, June, August, August, April, possibly Sept (there were various riots starting in 1991 and continuing sporadically until 1992 [Libcom]), May, August, and July.

Allowing an anomaly of October, these are all when the weather tends to be better. The recent riots stopped when the rain started. Just saying!

The list is incomplete, missing the Nottingham cheese riot from the start of October 1766 [Wikipedia-3]. Further worldwide riots are listed on another page [Wikipedia-4]. Try analyzing that if you have time on your hands.

Anyway, I am not an expert on socio-political matters but I know full well my sample size is too small to be statistically significant. If you find yourself noticing a pattern, do investigate, but hold a cynical thread in mind too, questioning your assumptions and analyses. Spotting non-existent patterns isn’t always problematic. Maybe you have looked at clouds and chatted with a friend about animals or other shapes you can see. This is an example of pareidolia, which Wikipedia [Wikipedia-5] describes as:

the tendency for perception to impose a meaningful interpretation on a nebulous stimulus, usually visual, so that one detects an object, pattern, or meaning where there is none

If you ever see the face of someone famous on your toast, you probably aren’t special.

A data visualization technique you might have come across before, uses faces to display data. These are called Chernoff faces ([Wikipedia-6] or the original [Chernoff73]), and use the size of various facial features, eyes, mouth, nose and so on, to graph features in data. You can easily use these for high dimensional data, and Chernoff used faces because people could spot similarities very easily. Our brains seemed to be wired to recognize facial features.

We’ve discussed pareidolia and people finding meaning where there is none, but we can equally miss meaning and reasons for events or outcomes where there is significance. For example, people might blame their compiler when their code doesn’t work. One case I came across involved a mysterious bug, where “code ended up in the wrong destructor”. After a few probing questions from Yours Truly, it became apparent the basics of virtual destructors and derived classes in C++ were unknown to the complainant. I am rather pleased with myself for managing to debug code without even seeing it. Now, it is easy to get deep into a problem though and miss what’s right under your nose, so maybe the other guy was tired and forgot the basics, so I might be jumping to conclusions. Nonetheless, if you go in with bias, you might fail to understand what is really happening or even see what you expect to see rather than what’s actually there.

Discussing a bug without the code in front of you can be a challenge. Furthermore, words are an imprecise way to communicate, which can make discussion difficult. Some of us are more precise than others. If asked, “Do you want tea or coffee?” some readers will say “Yes.” I guess you get used to how people tend to respond and might need to adapt your language to your target audience. Sometimes you think you are talking about the same thing, but later find a mismatch between the meanings or words or phrases. I may have mentioned before running a short coding practice sessions with colleagues, using the awesome cyberdojo [cyberdojo]. I paired up with a team mate, and she was very surprised to see me actually write a test first. I had said I like test-driven development several times, and used the phrase ‘test first’, but somehow those words meant something different to her. I learnt something and so did she, so learning as a team was a win. A much simpler example is someone going the wrong way. I told my Grandad to go left once, and he went right. The simplest solution to that problem was saying, “No, the other left.” He instantly turned the other way without arguing about the ridiculous phrase I had used. Words are weird.

We use precise notation for mathematics to pin down definitions exactly. However, such notation can be difficult to read, even if you do know some maths. One of my 78 browser tabs is a Nature article on fast matrix multiplication with reinforcement learning [Fawzi22]. It starts with an abstract, as most academic style papers do, talking about the use of machine learning for automatic algorithm discovery. All good so far. Then we get to the main article, which gives an overview of what’s coming up, as you would expect. Then the trouble starts. The first sentence tells me matrix multiplication is bilinear so can be represented by a 3D tensor and offers an illustration to help. It’s taken several attempts to figure out what the figure means. So, now I will have to go back to the text and try to read this again. Sometimes pinning things down precisely makes life harder, and yet it is good to be precise sometimes.

Even if you don’t need to be exact, knowing why things are the way they are, or a bit of history behind words or ideas can be useful, or at least interesting. Chris Oldwood shared some of his favourite quotes and aphorisms, back in 2023, and considered their origins [Oldwood23]. He began with the phrase “ship-shape and Bristol fashion” which sprouted a sidebar explaining in more detail, essentially everything being in the right place and able to deal with Bristol’s tidal river and lots of mud. Francis Glassborow has also been running a ‘Meaning of Words’ series in CVu, since 2021, with many disambiguating commonly muddled words like operator and function, code and cypher and others looking in detail at one or more ideas. I am on the verge of writing another book, and my new editor asked me what a compiler is. Fortunately, part 8 of Francis’ series is about ‘Assemblers, Translators and Compilers Revisited’ where he says [Glassborow22]:

Fundamentally, a simple compiler converts instructions written in a high level language (with a high level of abstraction from machine level code) into a lower level language with less abstraction.

Unfortunately, I would then have to define high and low level, and abstraction. Sometimes, you have a bootstrap problem. Do your boots have straps? Some, but not all, of mine do. There’s another odd phrase. Wikipedia [Wikipedia-7] suggests some possible etymologies, all referring to impossible tasks. Even if your boots have straps, you won’t be able to pull yourself out of trouble using them. But you might happily bootstrap a compiler. Try explaining that to someone who doesn’t know about computing.

The BBC news website, possibly inspired by Francis, has been running a series by ‘The Vocabularist’ for a while, and once attempted to explore the roots of the word ‘computer’ [BBC-2]. The article says ‘Computer’ comes from the Latin ‘putare’ which means both to think and to prune. It’s a jump to get from gardening to computing, but if you ever prune shrubs or tidy a garden, you are thinking and planning, and possibly even counting, buckets of rubbish, how many hours you have spent… Counting and computing seem to be related words. Of course, a computer is one who computes, so you have another bootstrap problem. How children ever learn language amazes me.

I would end by quoting Ecclesiastes; the second verse of the first chapter

“Meaningless! Meaningless!” says the Teacher. “Utterly meaningless! Everything is meaningless.”

However, that’s not a positive note to end on. Some things in themselves might be without meaning, but they can still be fun. You can describe music in technical terms, or via musical notation, but that doesn’t ascribe it meaning. It’s still fun, beautiful, inspiring or other word of your own choice. It’s easy to get distracted but allowing your mind to wander from time to time can be a good thing. You don’t need to ensure everything you do is purposeful, but do watch out for seeing patterns where there are none.

References

[BBC-1] ‘Why are there riots in the UK?’ posted 7 August 2024 and updated 9 August 2024 at https://www.bbc.co.uk/news/articles/ckg55we5n3xo

[BBC-2] The Vocabularist, ‘What’s the root of the word computer?’, posted 2 February 2016: https://www.bbc.co.uk/news/blogs-magazine-monitor-35428300

[Chernoff73] Herman Chernoff (1973) ‘The Use of Faces to Represent Points in K-Dimensional Space Graphically’ (PDF) Journal of the American Statistical Association 68 (342). American Statistical Association: 361–368

[Cyberdojo] Cyber-dojo: https://beta.cyber-dojo.org/creator/home

[Fawzi22] A. Fawzi, M. Balog, A. Huang et al. ‘Discovering faster matrix multiplication algorithms with reinforcement learning’ Nature 610, 47–53 (2022). https://doi.org/10.1038/s41586-022-05172-4

[Glassborow22] Francis Glassborow (2022) ‘Assemblers, Translators and Compilers Revisited’ in CVu 34.5, available at https://accu.org/journals/cvu/34/5/cvu34-5.pdf#Page=9 (You must be a member and logged in to see editions of CVu)

[Libcom] ‘Hot time: Summer on the estates – Riots in the UK 1991–2’ https://libcom.org/article/hot-time-summer-estates-riots-uk-1991-2

[Oldwood23] Chris Oldwood (2023) ‘Afterwood’ in Overload 31(175):19-20, June 2023. https://accu.org/journals/overload/31/175/overload175.pdf#page=21

[Wikipedia-1] ‘England riots’: https://en.wikipedia.org/wiki/England_riots

[Wikipedia-2] ‘Benford’s law’: https://en.wikipedia.org/wiki/Benford%27s_law

[Wikipedia-3] ‘Nottingham cheese riot’: https://en.wikipedia.org/wiki/Nottingham_cheese_riot

[Wikipedia-4] ‘List of riots’: https://en.wikipedia.org/wiki/List_of_riots

[Wikipedia-5] ‘Pareidolia’: https://en.wikipedia.org/wiki/Pareidolia

[Wikipedia-6] ‘Chernoff face’: https://en.wikipedia.org/wiki/Chernoff_face

[Wikipedia-7] ‘Bootstrapping’: https://en.wikipedia.org/wiki/Bootstrapping

Frances Buontempo has a BA in Maths + Philosophy, an MSc in Pure Maths and a PhD using AI and data mining. She's written a book about machine learning: Genetic Algorithms and Machine Learning for Programmers. She has been a programmer since the 90s, and learnt to program by reading the manual for her Dad’s BBC model B machine.