Why Radicals ARE NOT the "Building Blocks of Chinese Characters" 

These guys are my heroes. When you first start to study Chinese or Japanese, you make an innocent foray to the internet and you are unwittingly barraged with stupendous amounts of misinformation. I was fortunate enough to run across an Outlier blogpost about 3 months into my studies where they pointed out something like, “Radicals are a way to look up characters in paper dictionaries.” I will never forget that moment. After all these people had deigned to bestow upon me that radicals are the “key” to understanding characters. The bad news is that this nonsense is everywhere and most of the popular teaching venues are infected with it. The good news is that Outlier now commands a lot of respect and there are a growing number of resources that have followed their lead.

Thanks! I appreciate this!

When teaching pronunciation and reading, I like to use what I call character clusters to help my students remember. For example yán 言，严，盐，炎。The characters all have different meanings but have the same pronunciation.

Thanks for the kind words, Dr. Moser! Hope you're doing well!

🦀

either you are stupid or your joke isn't funny

Richard Spinoso We’ve got more coming! If you have any topics you’d like to hear about, or particular questions you may have, just let us know!

The radical only gives meaning about 64% of the time. That’s not enough to be useful-the remaining 36% means you definitely can’t just assume the radical has anything to do with the meaning of the character. Sure, the goal is to learn characters, not etymology. But learning the real reason a character is constructed the way it is will help you to learn the character. And not just that single character, but the whole writing system, because you’ll be able to see the real connections between characters that share components. Most sound components make a lot of sense, especially if you learn a few basic guidelines for understanding how sound components work. Welcome to the channel. You’ll find we’re focused on how explaining how Chinese characters *actually* work, and that learning the real story actually makes it much easier to learn characters!

1) "The fact that radicals are used to to organize dictionaries does not prove that radicals are not the building blocks of characters." That is true, but we're making no such claim. We're merely pointing out where the idea of radicals come from. 2) It's not about what it looks like. It's about facts. For instance, 丙's radical is 一. Early forms of 丙 don't even have that top stroke. It's what paleographers refer to as a 飾筆. So, 丙 is not "composed" of 一. the bottom part of 丙 is not a radical either. Even if you try and make the argument that 丙 is composed of 一+人/入+冂, you would not be able to understand the composition of 丙. The reason? It's not composed of any of those parts. The resemblance in the modern script is coincidental. So, 丙 is not composed of radicals. You also seem to be ignoring the fact that there is only one radical per character, which also means that characters are not composed of radicals. There are plenty of characters beyond the ones that have 1-6 radicals you're referring to (and by the way, the ones that have 1-6 as their radical are just like 丙 for the most part, i.e., completely random). But, you're making my point here. The considerations for deciding which part of a character is a radical are more broad than considerations about semantics. As such, it can't be true that radicals are simply semantic components. 3) No, 帛 is not the sound component in 錦. 金 doesn't give a meaning in 錦, and the meaning of 錦 is tightly related to the meaning of 帛. The sound of 金 is very close to the sound of 錦. As to 綿 and 棉, 綿 came first. Paleographers are not in agreement as to whether 帛 gives a sound in 綿. The initials m- and b- are in the same place of articulation (i.e., are closely related), but the main vowel is different. There is some uncertainty about the origins of the -n in mian. If it came from an -ng sound, then that would be compatible with the -k ending of 帛 (which has disappeared in modern Mandarin). Also, the 《說文》 analyzes it as 「緜(=綿)，聯微也。从系帛。」 The 从系帛 part is ambiguous. It's often interpreted as being an abbreviation for 从系帛聲, but others say it's 从系从帛. 棉 on the other hand, is derived from 綿. The 糹 was replaced with 木 to create 棉. If 帛 is sound component in 綿, then it is also sound component in 棉. If it's not, then the sound component in 棉 is 綿省聲 (many claims of 省聲 are not reliable, but in the case of 綿 & 棉 there is a very clear sound and meaning connection). 4) You make the claim "there is no such thing as a semantic component 'within' a character". Perhaps we mean different things. What I'm saying is that most characters have a component that indicates a range of possible meanings. I'm calling that component a "semantic component." You seem to be saying the same thing with your statement "The ‘semantic component’ of the character (usually the radical!) can give you a hint of the meaning," so I don't really know what to make of your claim that semantic components don't exist inside characters. This is quite a novel claim, so I'll leave it up to you to prove it. Also, it's not the character that gives meaning. It's that there is a spoken word, a pronunciation + meaning combination. It's the spoken word that has a meaning. The job of a character is not to create meaning, but to allow the reader to connect a spoken word (which has meaning) to a symbol on a page. It's just that most people say things like, "this character means....", but what they are really saying is "this character represents spoken word X, which means...." The proof is that the same character can represent multiple spoken words and the same spoken word can be represented by multiple characters. I'm not proposing that there is or should be a "perfect system." I'm saying that the idea that characters are made of radicals is one that is both not coherent with character formation and evolution (and thus not coherent with their current state), and it will make your learning more, not less, difficult. Understanding characters on their own terms, i.e., as combinations of components that point to the sounds and meanings of the words that the characters represent, is much easier. You don't have to deal with the many exceptions to the radical system that you briefly touched upon. A semantic component can always be understood as a semantic component. A sound component can always be understood as a sound component. Radicals on the other hand, are plagued with a huge amount of exceptions. The radical in a character basically gives no reliable information. About 70% of the time, it does correspond to a semantic component, but even if you know that, it says nothing about the rest of the character. This is borne out both by native speaker misunderstandings about characters, but to foreigners as well. For example, the radical of 祭 is 示. Knowing that does not elucidate the structure of 祭. Knowing that the top left is a piece of meat (i.e. a semantic component), the top right is a hand (another semantic component) and that the bottom is an altar (another semantic component) is helpful, especially when you know that it's a depiction of a hand putting a piece of meat on an altar to express the idea of sacrifice. There's only one radical here, which people who think of characters as being made up of radicals usually interpret as meaning "to show". Perfect knowledge of radicals gets you no closer to understanding character structure. Perfect knowledge of functional components gets you to being very clear on character structure, therefore I'm proposing that using functional components is better for understanding character structure, and therefore making learning easier.

@@ashhenson7214 I appreciate your response Ash. 1). I was saying that characters *can be broken down into radicals. Most characters can easily be divided into the radicals (7-214) with the exception of a small amount of irregular characters composed of nothing but strokes i.e. 乑, or a radical + decorative strokes i.e. 必, 甬 or 向. If it’s obvious the character has none of the radicals 7-214 it was most likely filed under radical 1-6. That’s how radicals and character indexing have always worked, and there are rules in which indexing radicals are allocated. 2). Re 丙; it doesn’t matter what the ancient and curvy character components were 1000 years ago. 丙 means “third of a kind”. You don’t need to keep the modern script’s character components semantically consistent with the ancient ones in order to know what this character means. So, 冂 is still a radical in modern script. 丙 is composed of 一+人+冂. (They’re not ‘semantic’ components but that’s irrelevant). It’s filed under the 一 radical because for 冂 to qualify as the indexing radical it would have to enclose all the character components, which is the case with all but two of the characters filed under 冂部(these are 冎 and 冓, but indexing these two under 冂 makes sense because using 口部 and 一部 would make counting the strokes and finding the characters more difficult.) 3)Re錦. 帛 is a sound component in 綿 and 棉. You don’t need to read up on history to ultimately pinpoint a sound component. Like you said, “palaeographers are not in agreement”, so we might as well just let it go. Simple fact is that 帛 does give the reader the sound clue in 棉 and 綿. It would be stupid for someone to make their pronunciations something like mu4 and mi4 respectively. And just because 帛 is pronounced bo2 that does not mean it’s sound cannot change when interpreted as one of those sound series. 錦 is an exception. There’s also not too many characters that use the indexing radical as the sound component. One exception like this does not trash the “radical + other stuff” generalisation, which does work with most characters. As you said, approx. 70%. As someone who learnt English as your first language (I presume) you’ll realise that our language probably has more phonetic inconsistencies than Mandarin. 70% is pretty decent if you ask me. And I don’t know why you disagree with me that 金 can’t be viewed as semantic component in 錦. Gold has been used in brocades for eons. But it still doesn’t matter. 錦 means brocade, that’s it. It really doesn’t help to divide a meaning into meanings. 4). Maybe I could have worded that “semantic comments” don’t exist” claim differently. All I mean is that it’s the character as a whole which gives the meaning. Debating what the semantic component is a waste of time and is just splitting hairs. E.g. When I learnt the word ‘autism’ in primary school my teacher didn’t break the word into ‘semantic components’ and tell me that I had to understand that auto(self) + -ism(way of thinking). It’s interesting, but it’s not needed. Autism means autism, that’s it. Same deal with characters. With the enormous pressure Chinese students see under they simply have no time to analyse character structure and evolution and mostly rely on the “radical + other stuff” thingy to help them recite pronunciations. Everything else just comes with practise. Lastly Ash, you said that Chinese characters don’t indicate meaning, they’re just “written symbols to represent a semantic sound”. Well, yes, but what exactly do you mean? I could say the same thing with any written language that has words. I don’t know why you said that the way you said it. Characters do indicate meaning, as all 8000 characters worth mentioning ‘have’ a meaning. 四 means four. Obviously it doesn’t mean legs in a box 口+儿. Who cares. With English, how do I know that ‘beautiful’ means beautiful? Who cares. I just learnt it. I’m sure if you stared at any English word for three hours straight you’d briefly get alexia and forget the purpose of the letters in the word! The written symbol would now mean nothing to you. That would happen with any language man. The character/semantic syllable (I don’t know what the perfect word or phrase is, even though everyone just calls them characters, and that’ll never change) both work together to link visual and auditory language. One syllable = one meaning; One character = one meaning. Character/syllable combined = one meaning That’s never changed, Chinese languages have always been this way, hence why the national language of China is Standard Chinese (not Mandarin).

@@mickyhermit4437 I think one of big difference here is that you're mixing surface structure and deep structure. Roughly speaking, surface structure is somewhat random (as your examples of radicals that have nothing to do with the character show). And indeed, radicals themselves are surface structure. The notion of "radical" didn't even exist for the first 1300 years of character formation and evolution. Deep structure is relevant to understanding a character's working parts and answers the question "why does this character look this way?" I'm not randomly making a criticism. Rather, I'm showing that radical + other stuff is a bad way of thinking about characters. If you don't like the phrasing "bad way", then replace it with "an inefficient way to learn characters that will have you constantly spinning your wheels." If you learn characters via functional components (i.e., if you understand the working parts), then you'll remember them more readily, recall them at a higher rate and be able to make intelligent predictions about characters you haven't learned yet. And, I don't mean after you've learned 1000s of characters. I mean very quickly. This claim is borne out by people that have taken our character course and use our character dictionary. "radical + other stuff" is an amalgamation of explaining things away, is weighed down with exceptions, etc. If you like that system, by all means, keep using it. But the claim in the video still holds: characters are not composed of radicals. There are many character components that cannot be broken down into radicals, and I'm not just talking about cases like 丙. If you're content with 70%, then more power to you. I'm saying you can know at much higher rate than that, closing in on 98%. Further more, I would argue that it's that 30% that's more important than the 70%. Such a high % of noise in your system is what causes all of the fuzziness. If you want clarity, we can give you that. I agree that students shouldn't be looking up ancient forms, meanings, etc. That's why we created a dictionary that does all of that for you based upon an actual understanding of the origin and evolution of each form (and the student doesn't even need to know any of that. They just need to know what the working parts are in the modern form and how that relates to the original meaning). I'm not sure where you get the notion of "One syllable = one meaning." I suppose if you add "at one time" or "in one case" it may work, but Mandarin is known for the crazy number of homonyms. "one character = one meaning" is equally odd to me. What do you make of the many senses given for characters in character dictionaries? Maybe you're just stating it in a way that doesn't make sense to me. "That's never changed". Well, if you want to keep that idea, you should probably not learn anything about ancient forms of Chinese, because they certainly don't hold that rule. Also "Standard Chinese" is exactly a form of Mandarin. When language reformers were making the push to get away from using Classical Chinese, their mantra was 我手寫我口. This is only true for Mandarin speakers, not for speakers of southern dialects (for instance). In standard Chinese, you write 他們想去哪裡? In spoken Cantonese, you'd say 佢哋想去邊度呀? The words for "they" (including the word for "he/she/it" and the suffix for plural) are different, the way you say "where" is different. Why? Because Standard Chinese is a form of Mandarin and is very different from dialects like Cantonese. Yet, Standard Chinese uses Mandarin forms. That's because it is Mandarin.

@@ashhenson7214 Alrighty...let’s get ready for round three. (I’m enjoying our discussion so far, just so you know.) I mean this sincerely, but you haven’t negated any of my points yet. All my point was that your blanket ban on “radicals being the building blocks of characters” and “sound components are terrible if there’s one exception like 錦” are just not factual. If you wanted to be neutral about it and suggest that there may be a cool trick that someone can use to learn a character better, you should have approached the issue like that (constructive feedback), not say ancient Chinese scholars are wrong, (who you’re referring to I don’t know) because they’re not wrong, and setting that kind of tone for your video will attract people who will want to debate. If there was a Chinese researcher who vehemently insisted that 帛 is the phonetic hint of 錦, then I say with confidence that he/she is not a Chinese character researcher. I’ve had a look, I cannot find any source that stated 帛 is the phonetic hint in 錦. The Shuowen dictionary and all modern sources I came across in my studies all say the same thing. All I see is things like, “in this case, the radical is the sound component’ blah and blah. Just because 錦 is jin1 that does not mean that 綿 and 棉 don’t use 帛as the sound component to give mian2, even thought 帛 as a standalone character is bo2. As I said earlier, the sound of 帛 changes when put into another character, but it’s still the sound component as these two characters have the same pronunciation. If this wasn’t the case, the pronunciation of 棉and 綿would NOT be exactly the same. I’m fully aware of deep structure, surface structure and all that mumbo jumbo but in academic Chinese language studies (not palaeography), there’s very little time to learn 4000-8000 characters and countless compounds words and idioms, (needed to be fully proficient, not just something like a HSK6). You can’t keep referring back to an ancient cursive script. The purpose of the development of standard Characters (正體字] was to eventually get rid of curved characters (e.g. seal script) and make the writing system more efficient. It’s a completely new system now. If your system thingamajig can teach you how to properly learn 8000 characters in 4 years I’d say good on you, simple fact is that many people have not used your system and still found “radical + other stuff” very useful (e.g. me and someone else who recently commented). It’s not a terrible/inefficient way to look at characters. That’s just your personal opinion. If you’ve helped people learn Characters from scratch, that’s great, I’m never said your work is wrong, but it simply does not disprove that radicals are the building blocks of characters as they can in fact be broken down. Re “radicals didn’t exist for x amount of time”. Seal script characters had no offical list of “basic building blocks”. If there was, there would have been reference ‘radicals’. But no dictionary like the Kangxi was compiled at the time, nor was it needed as writing was less prevalent. The curves were so inconsistent and kept on differentiated with so many characters. Modern characters 正體字 have much less variation. They could all be broken down into the shapes from radicals 7-214 or 1-6, (一丨丿丶乙 ) and radical #6 (亅部) as an extra reference. So radicals couldn’t exist for seal script. It was a disorganised system by today’s standards. I’ll bring back my original analogy that dictionaries need a reference particle to look up a word/character. In English, it’s A-Z. With Chinese, it’s radicals. Even though they’re literally called “section headers” in Chinese, again, EVERY character can be broken up into radicals 7-214 and 1-5 (一丨丿丶乙, the five basic strokes) and they added亅部 for good measure. (No examples of characters disprove this, including 丙 or 业 or whatever else there is). So radicals can be used both for dictionary reference AND breaking down characters, not just for dictionary lookup. But another thing you said which makes me wonder “why?” is your suggestion that Mandarin comes from Standard Chinese. Standard Chinese is a “language”, but it’s not a spoken one, or even a written one for that matter. All it means is that you can take characters/words from a number of Chinese dialects and pronounce it using the Chinese language of your choice. Hence the “standard” in Standard Chinese, and the “standard” in Standard Arabic. Cheeky example; 白鬼 is a derogatory Cantonese word for Caucasians. This is not a Mandarin word, but you can still pronounce it using Mandarin. If the Mandarin speaker also knows Cantonese, they will know what that word means. This is why the Chinese government could not get rid of characters. There were too many dialects to account for + cultural reasons and resistance. This is also why India has many official languages and not a national standard language. All the languages there have different writing systems, and hence why there is no possible way for there to be a language called “Standard Indian”. But with Chinese, a standard language is possible, because the dialects use characters. Hooray. Lastly, your other suggestion that one syllable ≠ one meaning, please give me an example of that. It’s definitely not like English, where you can have a five syllable word like “re-spon-si-bil-ity” with only one meaning. Chinese languages with characters are definitely 1 syllable = 1 meaning. They can be compounded to give new meanings, or borrowed purely for sound, but a character is still a character. 1 character = 1 meaning. You’ve also fallen into the common misconception that Mandarin is FULL of homonyms. People forget the tones. For it to be considered a homonym, it really should be a two character word with the same pronunciations AND the same tone, not just the same pronunciation with different tones. As you’ve aware, Mandarin is very tone sensitive. When you boil it down to that, the number of ‘homonyms’ decreases. But tones are not a part of IPA as I see it. So from an international phonetic standpoint you could call them homonyms. Who knows. Tones are not phonetic, they’re more like “harmonies”. This is why a lot of people think you can’t change your tone of voice in Mandarin.

@@mickyhermit4437 claim: "You don’t need to read up on history to ultimately pinpoint a sound component." You do when the sound component isn't obvious in Mandarin, like 各 in 路 for example. Sound components weren't created for Mandarin, so there are a good number of them that aren't recognizable based on Mandarin alone. claim: "Debating what the semantic component is a waste of time and is just splitting hairs." Then why are you debating it? You seem to be saying that not understanding character structure is better than understanding it. I'm not sure how you can defend that, especially in light of the fact that understanding the object of learning is the #1 rule for effective memory. claim: "Everything else just comes with practise." If there were true, then most people that studied Chinese would end up mastering the writing system. That simply isn't true. If it were, there wouldn't be an entire field dedicated to fixing or working around the problem. There are tons of books written to try and solve this exact problem. claim: "All my point was that your blanket ban on “radicals being the building blocks of characters” and “sound components are terrible if there’s one exception like 錦” are just not factual." I never said that sound components are bad. I would call that a gross misinterpretation of my views. To the contrary, they are very useful. And, if you know a few extra rules, they are extremely useful. That is a big part of why what I'm saying here is that understanding characters via the functional component framework actually makes sound and meaning representation in characters very clear. Thinking in terms of "radical + other stuff" actually hides sound and meaning connections in many cases. You're happy that radicals coincide with the semantic component 70% of the time and use 70% as a marker of success. The problem is, anyone looking at the sound and meaning of a character's components and comparing them to the sound and meaning of the character will, roughly speaking, get that 70% right, so you don't even need a framework for that. Where the framework becomes useful is explaining the other 30%, which “radical + other stuff” can't do. The functional component framework can explain ~98%. If you're happy with your 70% + memorizing a bunch of exceptions, then no need to look for a different framework (makes me wonder why you want to discuss these things). I'll rephrase my position: Not all characters can be decomposed into meaningful radicals. If you want to say that strokes are meaningful building blocks, then let's not even discuss it further. That would a pointless discussion. There are meaningful parts that make up characters that cannot be further broken down into radicals (without resorting to breaking them down into meaningless strokes). How do I know this? Because I've analyzed the most common 4000+ characters. In order to represent their meaningful parts on a computer, we've had to create workarounds for over 200 components that are not supported by Unicode. These components are not radicals (i.e., they are not on the list of 214 Kangxi radicals). So, because the list of 214 radicals does not contain all the meaningful components that make up characters, because most of the characters in use today were created before the notion of radicals even existed (and the majority of those still maintain their ancient structure), and because maintaining the idea that characters are made up of radicals necessarily entails having to resort to including meaningless strokes, my claim that “radicals are not the building blocks of characters” still holds.

Yes, that’s correct.

What he's saying is absolute major champions league BS.

@@mikedaniels3009 您既然如此说想必您另有高见，可愿分享？

It’s not exactly arbitrary, although it is for some characters. But a very large number of learners (and native speakers) believe that "radical = semantic component," which is not true and which we're trying to correct in this video. 聚 is not a 會意字, but a 形聲字. 取 is the sound component and doesn’t contribute to meaning at all in 聚. Sources: 李學勤《字源》 page 727 黃德寬《古文字譜系疏證》 page 1054 Even the Shuowen (說文解字) gets this one right: 从乑取聲 (meaning 乑 is semantic and 取 is phonetic). Page numbers here differ by edition. So, I'm not sure where you're getting your info on 聚, but whatever resource you're using, I'd recommend taking it with a large grain of salt. Chinese lesson for the day. ;)

I’m not sure what you mean about 矢 and 天, since they’re completely unrelated characters. But if you want to learn about the building blocks, I recommend checking out our free webinar (whenever you have an hour to spare): www.outlier-linguistics.com/pages/webinar-signup

i think you can't do it anymore. there has been changes for thousands of years we don't know for certain what comes from where. even professionals are still trying to find it out but there is no data anymore. i would think of it as collective expression of humanity rather than a simple reading system so don't try to think of it as a defined mechanical system.

Check out the character learning products on our site: www.outlier-linguistics.com/collections/chinese-characters/ Our focus is on teaching how to learn characters most effectively, using their real logic. I think you’ll like it!

Wendy Hong We use simplified characters here because most students are learning simplified, but our products all cover both simplified and traditional equally!

我們並沒有放棄正體字！只不過每次都要提供簡體正體兩種很累而已。

@wargent99 Please be respectful here. No need to be combative. Wendy asked a valid question. Traditional characters are not on average overly complicated, though there are a few that are. People from Mainland China may have problems with them, but that isn't due to structural problems. It has to do with the fact that they aren't trained in them. Taiwanese people have problems with Simplified as well, mostly for the same reason. It's a misnomer that Simplified characters are more simple. The "simplified" does not mean "made to be more simple." Rather, it means "made to have a smaller number of strokes." Even paleographer 裘錫圭 from FuDan university mentions in his book 《文字學概要》 that there are many simplified characters whose structures are broken due to the reduction of strokes.

....and of course the mayhem of having some characters simplified to (incidentally) more closely resemble one another doesn't support the "simplified is simpler" model. I learned traditional first (Taiwan/Hong Kong) and later had difficulty trudging through learning some characters all over again (also not fitting the model). All in all, it sure is useful to know simplified and traditional.....I don't know why anyone would NOT want to learn both. Castigating one or the other for any reason seems to simply go against language science. Anyways, really good video....

Jackie Otting Well, it’s hard to fit everything into one short video, but basically, “radical” is a poor translation of the word 部首. A better one would be “section heading.” Radicals were created as a means of organizing dictionaries. Paper dictionaries are traditionally organized into sections (部), with each character in a given section containing the same graphical component, which appears at the head (首) of the section. The radical is usually also a semantic component, but not always. That fact has led some people to call them “classifiers,” but this is a misnomer because radicals aren’t always semantic components. Sometimes they’re non-functional, as is the case with 耳 in 聚, and sometimes they’re sound components, as is the case with 錦, and so on. Unfortunately, that led to people using the word “radical” to mean “component,” hence the idea that there can be “phonetic radicals” and “semantic radicals.” But that’s just muddying the waters even further. Much better to use clear terminology-use “radical” when talking about dictionary organization, which is their purpose, and use “component” when talking about character structure. One issue with terms like “phonetic radical” is that it reinforces the incorrect idea that “radicals are the building blocks of characters.” People will often look for a list of radicals to learn, thinking they’ll learn all of the “building blocks” first. They’ll find the list of 214 Kangxi radicals, or 188 if they’re using the Oxford Concise English-Chinese Dictionary, or however many radicals are on the list they happen to find. They’ll soon learn that there are a lot of components not on the radicals list. This leads people to try to break down characters further and further into non-functional parts, because “characters are composed of radicals.” That’s why you see things like “偷 ‘to steal’ is a person 亻 sneaking into a house 亼 at night (when the moon 月 is out) with a knife 刂.” That may be a memorable story, but it’s not the character’s real structure, and it doesn’t work for characters like 喻 and 輸. We’ve written about this on our blog too, if you’re interested in further reading: www.outlier-linguistics.com/blogs/chinese/getting-radical-about-radicals It’s also worth pointing out that the Wikipedia entry for radicals agrees with us on this (we didn’t write it): en.m.wikipedia.org/wiki/Radical_(Chinese_characters)

I am thinking to develop one!

I would love to find this as well. This video just rocked my world as a Chinese beginner, because I thought the 214 radicals would be the cliff notes to finally making sense of the characters. But it sounds like Ash says there are many more possible components than the 214. So when you're learning a new Chinese word, how do you breakdown the character? Or do you just need a good teacher who can do that for you? Anyway, terrific video. I'm also checking out the recommended blog posts.

Im just learning chinese but, for what it's worth, why make the perfect be the enemy of the good? What if learning a list of the top 40 or top 200 radicals or components is "enough" to get going, and just keep in mind the "rules" about radicals arent absolute, as he stated. Just my advice:)

@@MrGottaQuestion Beautiful point! I really needed to hear that today, since I’m starting to burn out given my zeal for improving my Chinese reading comprehension.

You're looking for the book "Analysis of Chinese Characters". It isn't as useful as you might think.

he knocked this discussion out of the park!

If that's the case, then 上帝 help us.

Absolutely. Note that 錦 is the same in both simplified and traditional.

Ha ha. Yeah. I should have said, "Lead astray." I was having a hard time with all the people staring as they walked past!

Creative- Name The maximum number of radicals in a character is one. That’s part of the point: radicals are for dictionary lookup, not for character structure. A character’s radical is the component under which it’s listed in the dictionary. I’m not following how we “proved ourselves wrong.” The radical of 聚 is 耳, which has no function in 聚, and the radical of 錦 is 金, which is phonetic and not semantic. I think you’re using “radical” to mean “component.” That may be part of the confusion.

@@OutlierLinguistics Yeah, that's what I mean. Rewatching it, I thought the words radical and component are interchangeable. What I mean is that those components aren't exactly useless. In the examples, the components you provided in fact did have a use in understanding how to read a character. I think dictionaries are superfluous when it comes to Chinese writing because if you are educated in your components, you can get the meaning and the sound of the character just by looking at the components. However, you can't simply look at the phonetic or semantic component of a character and be able to grasp it in it's entirety by just that one component alone. Like vowels in the english language, components play off of each other and effect the usage of the other components in the character. To put the wrong component in a character would be akin to misspelling the word in English. What confuses me is the fact that some characters are made of less than two or more than two components. Logically, a character should be made of two to three components. One component denotes sound, the second denotes meaning, and the third could denote tone or some other pronunciation quirk. I chalk the fact that there isn't a uniform system up to the history of Chinese characters as a whole. The Qin supposedly unified the writing system, but they did not standardize the rules for how components work in characters. I believe it's because Chinese people learn characters in a rote memorization manner without looking at the deeper meaning of why Chinese characters are constructed the way they are, leading to the usages of components being lost on future generations. Not to mention that Hanzi are used by more than just Mandarin speakers, making understanding the construction of certain characters that much more difficult.

@@creative-name5279 You're right, they're not useless, and we never said they were. Our whole thing is trying to get people to learn characters in terms of their real structure. It's just that many people think "radical = semantic component," and this video is meant to show that that's not necessarily the case. There are a lot of common misunderstandings about radicals out there, some of which we've addressed here: www.outlier-linguistics.com/blogs/chinese/getting-radical-about-radicals I don't really think that characters "should logically be" any certain way though. That's just not how languages and writing systems develop, unless you're talking about conlangs, or writing systems like Hangul which were created rather than evolving. What about characters which just depicted something, like 人? Why does it need a sound component? People creating characters weren't thinking in terms of "one sound component, one meaning component." It was a lot more complex and organic (and over a long period of time). It was more along the lines of, "how do I represent this spoken word I'm trying to write down?" If it sounded really similar to another word, maybe they'd use that word's character and add a semantic component. But there are even some characters that are composed of two sound components and no semantic components, and of course a lot of characters with multiple semantic components. So I don't think the writing system is the way it is because Chinese people prefer rote memorization; rather, I think it is the way it is because it was developed organically by a lot of people over a broad geographical area and a long time span. Also, the Qin unification wasn't meant to be a way of standardizing all characters to a certain type of composition. It was about stamping out the forms of characters used in kingdoms other than Qin, and promoting the Qin forms as the standard. Fortunately, the Qin forms had a lot less corruption and were a lot more "orthodox" than the writing in the other kingdoms, but there was still some. And yes, understanding a non-Mandarin Sinitic language, or even Japanese or Korean, can often make it easier to understand the structure of characters. A lot of sound components that don't seem obvious from a Mandarin perspective will make a lot more sense if you know, say, Cantonese. Mandarin is the least conservative Chinese language in many ways.

@@OutlierLinguistics I see. So the Qin simply did away with any characters that were not Qin. What I thought the language reforms were were a compromise where characters created in each dialect that didn't have an equal character in the mandarin dialect were added to the dictionary. That would explain how dialects other than mandarin can understand Hanzi even though it was created for a mandarin speaking audience. Modern hanzi would have grammatical and phonetic components from all dialects across China, making it a mishmash of languages where to truly understand all of the characters, one must understand all of the dialects that contributed to hanzi. But this does not seem to be the case. It seems like to learn Hanzi, all you have to do is learn mandarin, then the composition of characters would fall into place. I assumed there would be a standardized composition as they already have a standardized sound system (pinyin) and standardized stroke order. I have a hard time grasping the organic nature of the language and how certain things simply don't make sense. I want everything to make sense, right down to the composition. Since government intervention was heavy in the creation of modern Hanzi, I assumed everything would be standardized, just like how we have a standardized Latin alphabet in the west.

@@OutlierLinguistics That's also the way I assume Chinese scholars created pinyin. They simply looked at the phonetic component(s) of each character and romanized them, that way people of the west could understand the sounds the characters make as well as a native Chinese speaker. A lot of people claim that the communist simplification program actually decreased literacy instead of increasing it as the party had claimed it would, which would mean the components that were changed in the characters were vital components of those characters. Then I looked at Japanese simplification and how they chose a homophonetic phonetic component that was simpler to write and replaced the more complex one. Since the only thing that changed was the phonetic component instead of the semantic one, literacy remained the same in Japan.

Actually, the only thing a radical can do reliably is to help you look up a character in a paper dictionary.

Scott Dunbar Do you have any evidence that gold or silver thread was used in brocades at the time the character 錦 was created? Do you have any evidence that the person who created 聚 was thinking about listening and that’s why he specifically chose a sound component containing 耳? Without evidence, it’s just conjecture, and you’re letting the radicals dictate things, which is backwards. The simple fact is, “radicals” were invented long after the writing system, as a means of organizing dictionaries, so radicals can’t possibly have been what people were thinking about when creating characters. 聚 being filed under 耳 is due to the fact that 取 isn’t a radical, and neither is 乑. That is, they filed it under the only option they had, regardless of the character’s actual structure. Plus, this isn’t “our analysis.” We base this stuff mostly on peer reviewed research by native speaking scholars in China and Taiwan. We’re just reporting the news, in other words.

@@OutlierLinguistics How can I trust a linguist who uses 'plus' as a clausal modifier or clausal conjunction in an English sentence?

@@OutlierLinguistics You're assuming that the explanation is to be found in diachronic linguistic analysis, when synchronic analysis might be more appropriate.

Atlanta Bill Using “plus” in that way is perfectly idiomatic. What’s the problem?

Atlanta Bill That’s a fairly abstract assertion. Can you get a bit more concrete? What’s the issue with our analysis in your opinion, and how would “synchronic analysis” improve our ability to explain these characters?

Functional components! Radicals are used to organize dictionaries. That's it. That's their entire function and reason for existence. The clue is in the name: 部首 literally means "section head." The radical is the character that appears at the head of a given section of the dictionary, and which all characters in that section contain. It usually is a semantic component in the character, but not always. For example, in 錦, the radical 金 is a sound component, and 帛 is the meaning component. In 變, the radical is 言, which is part of the sound component (攵 is the meaning component). Functional components are the building blocks of characters-they're what express sound and meaning. Not all functional components are radicals, and not all radicals have a function. This is why it's best to keep the two concepts separate in your mind. Hope this helps! You can read more here if you’re interested: www.outlier-linguistics.com/blogs/chinese/getting-radical-about-radicals www.outlier-linguistics.com/blogs/chinese/three-attributes-three-functions

@@OutlierLinguistics Thx for the speedy response :) So ur saying the components of characters are 'Functions', seems like it would make sense to say "forms" based on ur "three attributes of Functional Components vid #1" (just finished that vid right before ur response). It doesn't seem accurate at all that "Radicals are simply a dictionary organization tool" since the 1st documented dictionary came to being only during the seal script and many radicals can be traced back to even before the 1st recognized/documented script (Shang Oracle Bone). Radicals being a dictionary tool came much much later than their synthesis and the nature behind them. Every radical is a visual short-hand of forms that are by nature "picture like". 'Picture like' is comparable to saying 'pictorial'. All Chinese evolved from oracle script which directly evolved from the prehistoric art found earlier on pottery. Basically the Oracle script functioned as a 'pictorial shorthand' of their tradition of artistic visualization style. By nature they are still pictures, especially since that is the foundation of Chinese etymology and morphology. Your perspective / work is very thought provoking tho, so I appreciate that and I especially appreciate having the privilege of speaking to a Sino Linguistist! Def like u hear your reflections on my own. Much respect.

​@@tristate0mind Note that I'm not saying "the characters that serve as section heads didn't exist before dictionaries." What I'm saying is that their use as section headings ("radicals") doesn't (and can't possibly) pre-date dictionaries. The first dictionary (the 說文解字) is where the concept of "radical" (remember: "section head") comes from, because the dictionary was divided into (540) sections. Also, I'm not saying the components "are" functions, I'm saying they "have" functions. The function they have depends on which character they show up in. For example, 金 in 錦 is a sound component, while in 錢 it's a meaning component. In both, 金 is a radical, because both of those characters show up in the 金 section of the dictionary, regardless of the actual role of 金 vis-a-vis sound and meaning expression. "Functional component" and "radical" are roles that a given graphical element may or may not play in a certain context, *not* intrinsic, immutable qualities.

@@OutlierLinguistics "The characters that serve as section heads didn't exist before dictionaries." Confused whythat claim as the character's called radical's don't serve as a single function, since they play the role of semantic meaning(or hint), visual pictograph(the origin of their meaning[s], as well as phonetic sounds(or hint to it). The way I personally see them is like the atoms that make up [what is supposed to be] all characters, and the strokes are like the subatomic particles. By saying the radicals are just dictionary headers seems the equivalent to saying atoms are just periodic table markers, when the point is that the atoms [on the table] are the unit states which make up molecules [one scale down]. Everywhere else I've been studying from explains that these character's are called radicals because they are the "smallest forms / building blocks which all [other] characters are built from. If I understand correctly the term radical is referring to the smallest unit of meaning (aka morphemes), "roots of meaning in characters". Another definition I see it that "radicals are the smallest unit of meaning in written Chinese". What doesn't add up to either definition is that there are forms like "ㄚ" which one would think would be under the 214 modern radicals but it is only found with the list within this radical 濟 (this radical seems like it should be considered a combination of radicals rather than a radical). (whY isn't ㄚ a modern radical??? lol) When we consider that radicals contain inherent semantic meaning and many (if not all) serve as phonetic markers (if it's not all definitely like to know), to call radicals components, component functions, or forms doesn't seem to make a distinction compared to the existing term 'radical' (which refers to the morphemic nature of the form). There are no radicals that are not in origin pictographs 9def like to hear an example if this is untrue) and they still serve this original function, so it seems difficult to claim that radicals are not pictographs. The claim whole characters aren't pictographs seems like an arguable point, but to instead call characters 'pictographic ideograms' seems to be the most accurate description of them since all the building blocks of the ideograms are by nature pictographs. When the pictographs combine then they show ideas. If this is in any way inaccurate I'd definitely like clarification how exactly. Especially since this seems to be the over all message everywhere else I've seen so far. Thx for your reflections : )

@@tristate0mind You're conflating "radical" and "component" here. "Radicals" are section headings (hence the Chinese name 部首), while components are the building blocks of characters. The fact that some people (natives and non-natives alike) use the terms interchangeably doesn't make them actually interchangeable. We're arguing that the two concepts should indeed be kept separate because they're quite different. It seems like you haven't read the blog post about radicals that I linked to. I'd highly recommend it, as it explains our position on this matter pretty clearly. Not just our position, of course, but that of mainstream academia on this subject as well. Radicals and morphemes, or even components and morphemes, aren't analogous. Morphemes refer to speech. I can't agree whatsoever with the use of the term "ideogram" to refer to Chinese characters. This has been thoroughly debunked. Ideograms directly represent ideas without reference to language. Chinese characters directly represent language, and only represent meaning indirectly, via language. Logograms, yes, but not ideograms. They aren't "pictographs combine[d to] show ideas," and in fact very few characters in the modern script are composed this way. The vast majority are composed of one semantic component, and one phonetic component (chosen solely for its sound, and not for its meaning). If you're reading elsewhere that characters are pictures combined to form ideas, then I'd suggest that you're reading stuff by people who have no idea how the writing system really works. :)

Note that we didn’t say “radicals are the only way to look up characters in dictionaries.” ;)

@@OutlierLinguistics that's not what i meant as well, i mean for the most of native chinese speakers the radical is a way to search characters in the dictionary. but surely not for you to invent characters, at least modern Chinese speakers won't.

@@ryanlee5751 But why do you think I don’t know that? The first part is the point of the video (and we’ve written extensively on it elsewhere). The second part...we never implied that anyone should try inventing new characters.

@@OutlierLinguistics because the title use “building blocks ”to describe radicals , it does looks like the building blocks of Chinese characters but no one really does it that way . but sorry, I haven't watched the video 😅

@@ryanlee5751 The title is “Why Radicals ARE NOT the building blocks of Chinese characters.” Meaning that we’re explaining why they’re NOT the building blocks. 😉

The point is to not worry about radicals. Understanding which component is a radical (there is only one) is not helpful for understanding character structure. Each video is made to show a single point (in this case, characters are not made up of radicals). If you want to understand more, check out our blog www.outlier-linguistics.com/blogs/chinese. There's a post called Getting Radical about Radicals that goes into detail. Also, we have a webinar www.outlier-linguistics.com/pages/webinar-signup . During the QnA at the end, there's a segment about radicals.

The radical _may_ also be a semantic component, but it isn't _necessarily_ a semantic component. So it's not that it's "not at all" related to the meaning-it's that you can't _rely_ on it being related to the meaning. So it's much better to think of radicals as being for dictionary lookup only. That's what they were intended for, after all. I'd recommend this article if you want to learn more: www.outlier-linguistics.com/blogs/chinese/getting-radical-about-radicals

@@OutlierLinguistics Thank you, I really appreciate it

​@@OutlierLinguistics i have recently started to study chinese and have like 1k characters already under my belt. Of course i have noticed by now that these characters are made of the "radicals", but to be honest, i haven't really studied them and i am not familiar with their meaning. Should I go a step back and learn about the meaning of the radicals properly , or just skip this part?

@@richardhartung1576 Characters aren’t made of radicals though. That’s the whole point of this video. This blog post is worth reading: www.outlier-linguistics.com/blogs/chinese/getting-radical-about-radicals Characters are made of functional components: www.outlier-linguistics.com/blogs/chinese/three-attributes-three-functions

@@richardhartung1576 Basically radicals are just 1 part of the character. This isn't true for all characters (Example 涕) but usually if you join all the minor characters together you'll get something close to the meaning (Example 好).

The choice isn't between "radicals" and "rote learning." The choice is between "learning how characters ACTUALLY work" (what we teach), and "ignoring how characters work" (a category which both "learn characters via radicals" and "rote learning" fall under). We're not leading anyone astray, we're leading people toward an actual understanding of how the writing system really works. Radicals are simply a means of organizing and looking up characters in dictionaries. They aren't the "building blocks of characters." Functional components are.

Arrogance how? We're just sharing something the whole field of 文字學 knows, and most non-specialists don't.

James B The question you should be asking isn’t “how often is the radical also the sound component,” but how often is the radical NOT a semantic component. It could be semantic, sound, empty, or non-functional, but most people think radical = semantic component. And that’s just not a useful way of thinking about characters. Components should be thought of in terms of what function they have in a given character. Why add another layer of abstraction and talk about radicals? Our point is simply that you should use each concept for its intended purpose. Radicals are for dictionary lookup. If you try to think of them as semantic components, you’ll be right a decent amount of the time, but you’ll also be wrong fairly often. So why not just leave them for dictionary lookup, which is their purpose? And yes, the scholar was a very famous one, who many western students of Chinese would know. I knew who he was way before I went to grad school, or even moved to Taiwan. He’s also known as the “father of pinyin.” And his paper, and that very example, was quite widely cited by native speaking scholars.

@@OutlierLinguistics Zhou Youguang isn't exactly someone associated with character study though is he (pinyin though, sure). If he actually insisted the phonetic in 锦 was 帛 then yes that would be very silly, but are you sure his argument wasn't more just "The average reader could be forgiven for thinking that on the basis of wider generalities the character should be read like 帛, but of course if we all used pinyin instead the reading wouldn't be a problem at all"? That is, that he wasn't talking of the problem of characters in general? Personally I prefer works arranged by phonetic components, with the radical function demoted to the residue, the disambiguating afterthought, but as the phonetics are more numerous and complex, and possibly quite embedded, the look-up methods tend to remain frustratingly conventional. Point taken that the radical may not always be adding that much of a semantic element, but the student can't ever entirely blank the metal or ear or whatever (radical) just from the POV of componential recall, and often that may be all people are really meaning, especially when they're talking loosely. I wouldn't get too hung up on it.

James B No, his argument was essentially as it’s presented in the video, that sound components are basically “the part that isn’t a radical,” and that 帛 isn’t a good fit for 錦, and therefore sound components aren’t very reliable. Yes, it’s a terrible, absurd argument, but (probably partially since he’s a venerated scholar) it has been cited and propagated by numerous other scholars as evidence that the sound component system, and Chinese characters themselves, are broken. Works arranged by phonetic component can be useful, but they also tend to be unwieldy as you’ve observed. Not to mention you’re at the mercy of the editor’s understanding of paleography and phonology. 鄭張尚芳 is pretty reliable. Others.....not always.

@@OutlierLinguistics I just checked and 帛 is a radical in the Shuowen (number 283 according to chinaknowledge). Knowing a historical fact like that could only help, and might even have saved you the time it took the make this video, respond to comments, etc. Thanks for the 鄭張尚芳 pointer though, I came across him from the phonetic series on Wiktionary but usually just rely on Schuessler's Minimal Old Chinese thingummy. The more popular works in English thankfully don't make Father Pinyin's apparent "mistake" (I'd call it a mostly valid generalization actually) but like you say may need to be understood for what they are - often useful, or serviceable enough, sometimes less so.

@@jamesb5422 I'm quite familiar with the Shuowen. How is knowing that 帛 is a radical in the Shuowen meant to help? At most it shows that the choice of which radical a character gets filed under can vary from dictionary to dictionary, making it even more unreliable as an indicator of the component's function. The fact that a dictionary editor can choose to file a character under a different radical shows that 1) their purpose is dictionary lookup, 2) they have no place in a discussion about character structure, 3) a component being a radical is not an intrinsic property, but a choice made by a person. So the point is, when trying to understand character structure, why talk about radicals at all? You say they're "often useful, or serviceable enough, sometimes less so." So why not leave radicals for their intended use of dictionary lookup, and just talk about the actual function of the components when talking about character structure (the thing that is always useful)? Bringing radicals into a discussion about character structure is like saying bright red cars are fast. Sure, it may be the case that most bright red cars do happen to be sports cars, so it's a "serviceable enough" indicator that the car may be fast, but the cars aren't fast because somebody decided to paint them red. We're suggesting that people should talk about things like torque and horsepower (component function) rather than paint (which section of the dictionary the character happens to be filed in), and you're saying "but paint works sometimes!"