Random Stuff

Good trivia questions are head-final, as opposed to head-initial. Practically speaking, this means that the answer to the question should ideally be narrowed down as much as possible as soon as possible. Good trivia questions can be hard to come up with because English is considered a naturally head-initial language. Consider the following questions:

• The Republic of China, with its capital in Taipei, is more commonly referred to by what name? (answer: Taiwan)
• What do people usually call the place that officially goes by the name "The Republic of China"? (answer: Taiwan)

In the first case, the immediate mention of "The Republic of China" is rather unusual; if we were asking something very specific about Taiwan, we would expect the questioner to use the word "Taiwan" as opposed to adding a needless layer of obfuscation. So an experienced answerer would expect that, if "The Republic of China" were replaced by "The People's Republic of China", then the answer would be obvious. So practically speaking, the most likely things they could ask are "give the common name" or "give the capital", or maybe "give the currency" or "give the ruling party". In any case, the first 4 words narrow down the range of reasonable answers significantly.

On the other hand, in the second case, we have no idea what universe we are even operating in before the last 4 words of the question.

Good trivia questions should be answerable by those who don't exactly know the answer, but are knowledgeable enough about the topic of the question that they can make an educated guess. Consider the following two questions:

• 500k or 5 million - which number is closer to the current population of Sydney? (answer: 5 million)
• 4 million or 5 million - which number is closer to the current population of Sydney? (answer: 5 million)

The first question rewards those who have some sense of what the population of a metropolitan city should be. The second question rewards very specific, non-transferrable knowledge about the population of a specific city.

The range of possible answers should get narrower (which is better) or stay the same (which is acceptable) as the question goes on, based on the most reasonable interpretation of the question up to any given point. But it should never expand or jump to a completely disjoint set of possible answers. Consider the following questions:

• If measured from the sea floor as opposed to from sea level, which mountain would be considered the tallest on Earth? (answer: Mauna Kea)
• Which mountain is the tallest on Earth, if we measure not from sea level, but rather from the sea floor? (answer: Mauna Kea)

In common usage, the term "tallest mountain" itself implies that we are measuring from sea level. In the second question, the possible range of answers changes from just Mount Everest, to a list of mountains that excludes Mount Everest. This punishes quick thinkers.

Consider the following question:

• Pyotr Ilyich Tchaikovsky composed an overture celebrating which country's victory over Napoleon? (answer: Russia)

Even if you know nothing about the Napoleonic wars, if you have the inuition that "Pyotr Ilyich Tchaikovsky" is an extremely Russian sounding name, then you would naturally guess "Russia" for this question.

Broad knowledge of linguistics, specifically of romanization, phonotactics and cultural stuff, can be used to make very accurate educated guesses in many situations. The questioner could either lean into the idea of this skill being universally applicable, or try their best to avoid situations where answers could be "cheaply" guessed in this manner. However, the questioner should not try to subvert this skill. For example, the question

• Hausmann was commissioned to renovate which European city? (answer: Paris)

may be considered be a bad question, since Hausmann is a relatively obscure figure, and all educated guesses based on the very German-sounding name will lead you astray.

The heart of this entire topic is the idea that not all chord progressions are made equal. That may sound obvious, but in reality few people have an intuitive grasp on this idea. Some progressions like bVI-bVII-I are inherently cadential. Others like iii-IV-V-vi are inherently transitional. Some like VI-V-vi-vi work very well with repeating 3-note melodies. Others like ii-V-I-VI, not so much. Now to begin, I claim that the following composition workflow is bad.

1. Have an idea of a melody in your head.
2. Come up with a chord progression, choosing chords where the melody notes create nice harmonies with the chord tones.
3. Change up the melody to be more interesting.

Conversely, the following workflow is superior.

1. Decide which family of chord progressions you will focus on.
2. Create a melody, using techniques you specifically know work uniquely well for such chord progressions.
3. Reharmonize.

Why? Simply put, the first workflow sort of presupposes that all chord progressions are basically the same, and that if chords locally harmonize well with the melody, then that sequence of chords will form a decent progression. Conversely, the second workflow segregates your knowledge of music theory into different harmonic contexts.