Want to know more about those inscrutable image-only radicals? (Like how to represent them in text?) While most are indeed “wanikani-only” radicals (not part of 214) they are to a large extent genuine components with listed code points in unicode, many of which have associated meaning/phonetic value. Some of these components are characters of their own in Chinese but either don’t exist or are extremely rare in Kanji. Only a few wanikani radicals do not have a faithful unicode parallel, usually because they consist of multiple components which are not grouped together as one component, as say, 殳 is. These don’t have meanings or readings beyond the sum of their parts, or-- and this is extremely unlikely-- they do and have been overlooked in the 87,882 ideographs in the CJK block. Conversely, some on the other hand represent similar-looking parts of different components (with different meanings) that don’t exist in isolation, let alone as a unified component. And finally, a few are total bullshit, IM TALKING ABOUT YOU LEAF. So, here we go, radicals and “radicals” with their legit characters in roughly level order. Some of these you will be very familiar with and have probably thought “why the fuck is this an image” and others you may, like I have several times in the last few hours, think “holy fuck this is a thing?”
Some of these might not render correctly depending on your unicode coverage (I’m talking about Creeper, Pope, and Egg) but you can still c+p them into the google and look em up. Items with * can be typed more or less, but aren’t really “things”-- see the notes
Gun 𠂉
Leaf 丆*
Stick 丨
Triceratops ⺌
Hat 𠆢
Beggar 丂*
Viking 龸
Kick 𧘇
Horns 丷
Spikes 业
Water Slide ⻌
Cloak 𠃌
Cleat ⺤
Train 𡗗
Pope 𥃭
Chinese 𦰩
Bear 㠯
Boob 𠫓
Blackjack 龷
Creeper 𠮛
Bar 㦮
Grass ⺍
Saw 巩
Squid 㑒
Zombie 袁
Death Star 俞
Morning 𠦝
Egg 𠀉
Shark 烝
Satellite 䍃 (I guess the image we see is a variant of 䍃 not given a separate point, somehow wikipedia renders it correctly in the infobox but not the title, at least for me)
＊Notes:
First off, this notes section opens a door to the harsh realization that kanji composition isn’t as easy as WK makes it seem. To start with, kanji can only have one radical, although they may have many components. Sometimes, after the components are accounted for, they have some extra strokes. These extra strokes can often be represented by components and combinations of components, even though the components aren’t actually in there.
Leaf:
Is bullshit. The image resembles 丆 (formed of the one radical + a stroke), a Chinese character which is not a legitimate component (it has no descendants, aka characters with it inside them) but could be used to represent a pattern of strokes. But the 一+丿 shape doesn’t appear in the kanji leaf is supposedly in. It seems to subsume every horizontal line connected to something below it.
百 画 夏 再 乏
Most egregious is 百, because it’s a 白＋one stroke. WK could have easily made it 白＋一, it’s actual components.
(Note that 一 here refers to a horizontal stroke, it doesn’t mean one and it’s basically impossible to tell when it does without knowing the history of the ideograph. But that’s the stuff I mentioned WK ignores and you should ignore it too because it’s fucking stupid.)
Beggar:
Is not a “thing”, because it is five things:
- The kanji 与 without the line in the middle (与、写) < the shinjitai are related but the original 與 and 寫 aren’t
- The phonetic component 丂 (巧、拷、考)
- The bottom of the phonetic component 夸 (誇、汚) < simplified to 亐 in the latter
- A thing in 極
- 亏 without a line on top (鰐)
Now, what you’ve all been waiting for…THE UNTYPABLE! (research ongoing)
Hills:
A completely unnecessary reduction of the kanji/component 之, taking away the dot, which happens to be the radical in 之. All three example kanji (乏、之、芝) have 之 clearly visible, so why reduce it? Answer- to protect the precious “leaf” in 乏. They should really just represent this with 之, which is actually a thing, unlike hills (which isn’t in unicode because it’s not a thing).
Hick:
Hick should really be in unicode, because it’s likely a phonetic component-- 度、庶、and 席 have very similar readings. I’m looking for it right now.
Cactus:
I’m almost positive this is in unicode, ima keep looking for it.
Tofu:
Tofu is a similar looking part of three components, two of which are related.
- 乑 without the two dots on the left (衆)
- The common element in components 旅 (旅) (and 𠂢 (脈、派)
There should be and might be a unicode to cover the latter, or at least 旅/祣 without 方/衤 on the left.
Explosion:
Explosion is shinjitai used for the bottom two forms of a tripled component, when those bottom components aren’t omitted entirely (if doing so doesn’t make the kanji look like another). I think that this only occurs when there is another component above or below the triple one. This doesn’t happen in rarer kanji though and for 品 (which can be simplified so the two mouths on the bottom are connected.) I’m looking for it because it really should exist.
Worm:
Not a thing.
-comb
-potato
-trident
-corn
-elf
IN PROGRESS
And as a bonus,
Legit components represented by bullshit unicode.
Lion (L) is 𠃊/𠄌
Pelican (ネ) is 礻/ 衤(示/衣)
Note: I’m not saying the names are correct (they aren’t) or that the images are perfect representations of the characters (some aren’t). Tell me if I’ve missed something, if you guys find it helpful (and help 
) I’ll add “Legit components represented by half-bullshit unicode” (where multiple similar-looking components are represented by one component).
