1 00:00:01,920 --> 00:00:06,560 Speaker 1: Welcome to brain Stuff production of I Heart Radio. Hey 2 00:00:06,600 --> 00:00:11,000 Speaker 1: brain Stuff Lauren Bogelbaum. Here in the movie The Sunshine Boys, 3 00:00:11,119 --> 00:00:14,640 Speaker 1: an aging vaudeville comedian explains a classic truism of comedy 4 00:00:14,640 --> 00:00:18,599 Speaker 1: to his nephew, The k sound is always funny. The 5 00:00:18,680 --> 00:00:22,239 Speaker 1: comedian played by Walter Matthau said, fifty seven years in 6 00:00:22,239 --> 00:00:24,560 Speaker 1: this business, you'll learn a few things. You know what 7 00:00:24,720 --> 00:00:27,400 Speaker 1: words are funny and which words are not funny. Alka 8 00:00:27,480 --> 00:00:30,280 Speaker 1: Seltzer is funny. You say alka Seltzer, you get a laugh. 9 00:00:30,720 --> 00:00:34,479 Speaker 1: Casey Stengel, that's a funny name. Robert Taylor is not funny, 10 00:00:34,840 --> 00:00:38,440 Speaker 1: Cupcake is funny, Tomato is not funny, Cleveland is funny, 11 00:00:38,720 --> 00:00:42,200 Speaker 1: Maryland is not funny. Then there's chicken. Chicken is funny, 12 00:00:42,560 --> 00:00:46,040 Speaker 1: Pickle is funny. And it's true. If you need a 13 00:00:46,040 --> 00:00:48,400 Speaker 1: place name for a punchline, you're guaranteed to kill with 14 00:00:48,479 --> 00:00:53,680 Speaker 1: Kalamazoo's connecticy or Rancho cucamonga. But why? Psychology professor Chris 15 00:00:53,680 --> 00:00:56,720 Speaker 1: Westbury at the University of Alberta has a fascinating theory, 16 00:00:57,000 --> 00:01:00,120 Speaker 1: and it's based on perhaps the two unfunniest words in 17 00:01:00,160 --> 00:01:05,319 Speaker 1: the English language. Statistical probability Westbury published a paper in 18 00:01:05,360 --> 00:01:08,880 Speaker 1: October of eighteen in the Journal of Experimental Psychology with 19 00:01:08,920 --> 00:01:13,120 Speaker 1: a first rate title, Reriggly, squiffy, lemmicks and boobs. What 20 00:01:13,319 --> 00:01:16,440 Speaker 1: makes Some Words Funny? In his research, he started with 21 00:01:16,480 --> 00:01:19,080 Speaker 1: a list of the five thousand English words rated funniest 22 00:01:19,120 --> 00:01:22,320 Speaker 1: by real humans and constructed a working mathematical model for 23 00:01:22,440 --> 00:01:25,679 Speaker 1: predicting the laugh factor of nearly every word in the dictionary. 24 00:01:26,480 --> 00:01:28,840 Speaker 1: When Westbury applied his model to a data set of 25 00:01:28,880 --> 00:01:32,399 Speaker 1: forty five thousand, five hundred and sixteen English words, it 26 00:01:32,480 --> 00:01:34,840 Speaker 1: decided that these ten words were the funniest of all. 27 00:01:35,319 --> 00:01:40,800 Speaker 1: Up chuck, bubbly, boff, wriggly, yeaps, giggle, cooch, gafa, puff ball, 28 00:01:41,000 --> 00:01:45,600 Speaker 1: and giggily. Runners up included squiffy, flappy, and bucko, and 29 00:01:45,720 --> 00:01:48,000 Speaker 1: the perennial favorites of every eight year old on the 30 00:01:48,040 --> 00:01:51,560 Speaker 1: planet poop, puke, and boobs. On the other end of 31 00:01:51,560 --> 00:01:54,000 Speaker 1: the spectrum, the word found to be the absolute least 32 00:01:54,000 --> 00:01:59,800 Speaker 1: funny was harassment. In his paper, Westbury explains that philosophers 33 00:01:59,800 --> 00:02:03,000 Speaker 1: have been trying to unraffle the mystery of humor for millennia. 34 00:02:03,240 --> 00:02:06,360 Speaker 1: Plato and Aristotle weren't big fans of humor, seeing it 35 00:02:06,400 --> 00:02:09,600 Speaker 1: mostly is a way of denigrating and feeling superior to others. 36 00:02:10,560 --> 00:02:14,760 Speaker 1: Casaro introduced incongruity theory, writing that the most common kind 37 00:02:14,760 --> 00:02:17,239 Speaker 1: of joke is when we expect one thing and another 38 00:02:17,360 --> 00:02:20,840 Speaker 1: is said, in which case our own disappointed expectation makes 39 00:02:20,919 --> 00:02:25,120 Speaker 1: us laugh. While the incongruity theory of comedy makes perfect 40 00:02:25,120 --> 00:02:29,720 Speaker 1: sense of even orangutans find switchery tricks hi hilarious, Westbury 41 00:02:29,760 --> 00:02:32,280 Speaker 1: says that it's not a true scientific theory and that 42 00:02:32,400 --> 00:02:35,200 Speaker 1: clearly not every incongruous event is as funny as another. 43 00:02:35,800 --> 00:02:38,600 Speaker 1: A random coughing fit in a crowded movie theater isn't 44 00:02:38,639 --> 00:02:41,480 Speaker 1: nearly as comical as a random farting fit. I mean, 45 00:02:41,720 --> 00:02:44,919 Speaker 1: just try to say random farting fit without smiling. So 46 00:02:45,160 --> 00:02:48,239 Speaker 1: the goal of Westbury's modeling experiments was to go beyond 47 00:02:48,280 --> 00:02:52,160 Speaker 1: philosophical theorizing and come up with a truly quantifiable scale 48 00:02:52,200 --> 00:02:56,080 Speaker 1: of funny. To do it, Westbury analyzed words in two 49 00:02:56,120 --> 00:02:59,600 Speaker 1: different ways, by their meaning and by their form. For 50 00:02:59,680 --> 00:03:03,000 Speaker 1: the first analysis, the researchers looked at semantic predictors that 51 00:03:03,200 --> 00:03:06,800 Speaker 1: group words with similar meanings using a free tool developed 52 00:03:06,800 --> 00:03:09,360 Speaker 1: by Google that identifies words that are commonly used for 53 00:03:09,400 --> 00:03:13,040 Speaker 1: one another a k A co occurrence. Westberry mapped out 54 00:03:13,080 --> 00:03:16,400 Speaker 1: the semantic relationships between two hundred and thirty four of 55 00:03:16,440 --> 00:03:20,840 Speaker 1: the human picked funniest words. From this correlation plot, the 56 00:03:20,880 --> 00:03:29,600 Speaker 1: researchers identified six different clusters or categories of funny words insult, sex, party, animal, bodily, function, 57 00:03:29,840 --> 00:03:34,840 Speaker 1: and expletive. Now this is where things get dangerously mathematical. 58 00:03:35,600 --> 00:03:37,840 Speaker 1: Since many of the words on the Human Rated funny 59 00:03:37,880 --> 00:03:41,040 Speaker 1: list fell into more than one category, the researchers needed 60 00:03:41,080 --> 00:03:44,000 Speaker 1: a more precise measurement of how a words meaning translated 61 00:03:44,040 --> 00:03:47,400 Speaker 1: into comedy. Using the Google tool, they came up with 62 00:03:47,440 --> 00:03:50,120 Speaker 1: lists of words most closely related to each of the 63 00:03:50,120 --> 00:03:53,320 Speaker 1: six categories, and then they came up with the average 64 00:03:53,400 --> 00:03:56,720 Speaker 1: values for each of those word categories using something called 65 00:03:56,720 --> 00:04:02,760 Speaker 1: linear regression analysis. Those average values for each category you know, insult, explicit, etcetera. 66 00:04:03,000 --> 00:04:08,880 Speaker 1: Became known as category defining vectors. When looking specifically at meaning, 67 00:04:09,120 --> 00:04:11,920 Speaker 1: it turns out that the funniest words don't necessarily fall 68 00:04:12,040 --> 00:04:15,600 Speaker 1: cleanly into the most categories, but are the words whose 69 00:04:15,640 --> 00:04:19,440 Speaker 1: mathematical values are the closest on average, to those six 70 00:04:19,600 --> 00:04:23,560 Speaker 1: category defining vectors. Here's how Westbury summed it up in 71 00:04:23,560 --> 00:04:26,920 Speaker 1: a press brief, The average similarity of a words meaning 72 00:04:27,040 --> 00:04:30,279 Speaker 1: to these six categories is itself the best measure we 73 00:04:30,360 --> 00:04:33,400 Speaker 1: found of a word's funniness, especially at the word also 74 00:04:33,480 --> 00:04:38,080 Speaker 1: has strongly positive emotional connotations. But meaning is only one 75 00:04:38,120 --> 00:04:41,039 Speaker 1: type of measurement. Westbury and his team looked at the 76 00:04:41,080 --> 00:04:43,800 Speaker 1: form of funny words, things like word length or the 77 00:04:43,839 --> 00:04:47,320 Speaker 1: individual sounds or phonemes that make up each word. In 78 00:04:47,360 --> 00:04:50,560 Speaker 1: the second analysis, the data fit nicely with the incongruity 79 00:04:50,600 --> 00:04:53,640 Speaker 1: theory of humor. It turns out that the fewer times 80 00:04:53,640 --> 00:04:56,200 Speaker 1: a word or its phonemes appear in the language, the 81 00:04:56,200 --> 00:04:59,400 Speaker 1: funnier we think they are. That helps explain why there 82 00:04:59,440 --> 00:05:02,839 Speaker 1: are so many k and oo sounds in funny word lists. 83 00:05:02,880 --> 00:05:06,279 Speaker 1: They're statistically improbable, or it's ending in l e like 84 00:05:06,440 --> 00:05:09,720 Speaker 1: wattle and wriggle, or another source of funds suggesting, as 85 00:05:09,760 --> 00:05:13,480 Speaker 1: the study put it, repetition, usually with a diminutive aspect. 86 00:05:14,560 --> 00:05:17,880 Speaker 1: So why are we laughing? Now? This is where things 87 00:05:17,920 --> 00:05:21,279 Speaker 1: get really weird. The human brain, it seems, is running 88 00:05:21,320 --> 00:05:24,520 Speaker 1: all of these complex mathematical models all the time without 89 00:05:24,560 --> 00:05:27,280 Speaker 1: any of us knowing it. As we watch TV and 90 00:05:27,320 --> 00:05:30,279 Speaker 1: read and talk to people. Our brains are constantly parsing 91 00:05:30,360 --> 00:05:34,839 Speaker 1: language for subtle semantic cross connections and statistical probabilities, and 92 00:05:34,880 --> 00:05:37,560 Speaker 1: the result, at least on this basic one word level, 93 00:05:37,920 --> 00:05:41,960 Speaker 1: is what we call humor. Westbury said, if asked which 94 00:05:42,080 --> 00:05:44,479 Speaker 1: letter is more common P or B, I think the 95 00:05:44,520 --> 00:05:48,320 Speaker 1: average person would have no clue consciously, but unconsciously they 96 00:05:48,320 --> 00:05:50,680 Speaker 1: are sensitive to that. And we know that because their 97 00:05:50,680 --> 00:05:54,960 Speaker 1: funniness judgments are reflecting exactly that kind of fine tuned calculation. 98 00:05:55,600 --> 00:05:58,719 Speaker 1: In other words, said Westbury, people are using emotions to 99 00:05:58,760 --> 00:06:03,160 Speaker 1: do math. Westbury argues that all of this makes perfect 100 00:06:03,160 --> 00:06:06,880 Speaker 1: sense evolutionarily, our brains have been hardwired over millions of 101 00:06:06,960 --> 00:06:09,440 Speaker 1: years to identify anything that's out of the ordinary as 102 00:06:09,440 --> 00:06:13,640 Speaker 1: a potential threat, and human emotions, including humor, likely developed 103 00:06:13,680 --> 00:06:17,280 Speaker 1: his ways of responding to improbable events and environments. O. 104 00:06:17,400 --> 00:06:21,120 Speaker 1: Westbury summed it up, people laugh based on how improbable 105 00:06:21,279 --> 00:06:25,320 Speaker 1: the world is. Of course, it's a long conceptual leap 106 00:06:25,360 --> 00:06:28,560 Speaker 1: from predicting the funniness level of individual words to modeling 107 00:06:28,560 --> 00:06:30,719 Speaker 1: the communic mechanics of a knock knock joke or a 108 00:06:30,800 --> 00:06:35,159 Speaker 1: salty limerick. But Westbury's work points the way maybe someday 109 00:06:35,160 --> 00:06:38,960 Speaker 1: we'll finally understand why that chicken crossed the road. One 110 00:06:39,000 --> 00:06:41,520 Speaker 1: thing is clear, though a frog wouldn't have been half 111 00:06:41,520 --> 00:06:48,960 Speaker 1: as funny. Today's episode was written by Dave Ruse and 112 00:06:49,000 --> 00:06:51,440 Speaker 1: produced by Tyler Playing. For more on this and lots 113 00:06:51,440 --> 00:06:54,080 Speaker 1: of other flappy boff topics, visit our home planet how 114 00:06:54,120 --> 00:06:57,039 Speaker 1: stuff works dot com. Brain Stuff is production of I 115 00:06:57,120 --> 00:06:59,479 Speaker 1: Heart Radio. For more podcasts, to my Heart Radio is 116 00:06:59,520 --> 00:07:02,240 Speaker 1: the heart dio app. Apple podcasts are wherever you listen 117 00:07:02,279 --> 00:07:03,159 Speaker 1: to your favorite shows.