The Structure of Moe

Or, Whence the Urge to Burn and Protect?

I’ve been having odd thoughts lately, mostly when I walk to and from class — but also in the shower (both places from which ideas emerge).  Where does moe come from?  That’s the question underlying our work here today.  I’m not going to quibble about the definitions of what moe is, I’m going to try to examine where it comes from.

Moe is typically viewed as a structural element.   Simply, fans view moe as something in the text that they decode.  It’s an emotional reaction fans have with the text, but the beginnings of moe itself are within the text.  To be a little more precise, the text does something, performs some action or makes some reference (whatever it is we view as moe), and we read it there and respond appropriately, according to our interests.

Here’s my thought:  moe isn’t a structural element; it’s a phenomenological element in the space around the text.  That is, we read into a text the moe we feel, rather than read from a text the moe we feel.

This construction might sound like it’s splitting hairs, but the implications of each view are very different.  If we view moe as structural and a constituent part of the text then we must feel the moe to read the text.  Besides being authoritarian, this stricture is also theoretically problematic.  If one fan does not see the moe that is, apparently, inherent in the text, that fan has actually not read the text.  This is different than seeing it and not enjoying it.

Consider horror, an entire genre based (generally) on the emotional response of the reader.  Can we say “horror” (however we define it) is structural?  I think so.  We can point to the elements of horror that always happen in texts (or almost always), even if we don’t feel any fear or disgust ourselves.  Some examples, taken at random, would be the attempts to undermine the typical societal view of our own well-being or strength; the highlighting of the horror of birthing; or the horror of the body (check out I Am Legend, Rosemary’s Baby, and The Three Imposters respectively).  We may not feel the emotional, phenomenological aspect of the horror, but we can see horrifying elements within the text.  To not feel horror is not to mis-read the text, but to not see the undermining elements is.

Back to moe.  Let’s pull out an example of moe.  Pontifus has some interesting examples here.  The urge to protect these girls, if in the text — that is, structural — means it is equally “solid” within the confines of the text as the guns and the music.

Can moe be, instead, phenomenological?  Might we consider it a reaction within the fans, or the fan-group, and not something woven into the narrative, imagery, &c.?

I think we can.  I am not suggesting moe is entirely woven in the fan-space, like many slash relationships.  Certainly there are typically markers in the text on which moe is built, but those markers are not, in themselves, moe.  We have coded them as such in the fan-space, the viewing gestalt.  Hence the arguments as to the definition of moe.  We cannot define something concretely that is entirely phenomenological; in turn, we cannot insist on readings that deal with moe structurally.

[written during a class's library instruction period and the break immediately afterward -- in short, sorry for the short post]

About these ads
Leave a comment


  1. *cough*

    Hiroki Azuma separated (as far as I can deduce from the skimming I’ve done of Otaku) the concept of moe from what he termed “moe-elements”: moe was the feeling attributed to characters, moe-elements become elements abstracted from those characters that one has felt “moe” for and entered into his “database” concept. The easy, quick example is “twintails -> tsundere” where “twintails” become a moe-element corresponding with the “tsundere” personality (I could go on and talk about the tsundere personality itself being a moe-element for a more fleshed-out personaity which is itself…you get the idea and I’m going to save you the verbiage and headaches). In a word: fetishization (which we all kind of know about, except not).

    I should note that Azuma implicitly suggests that his theories about moe and moe-elements is not limited to otaku anime or even anime in general; rather, he’s using them as an example to form an image of the reading habits of the postmodern world (where meaning imposed by an author and/or society is absent, and the reader is left with a self-constructed “database” of narrative codes). I also like to add my own caveat to his ideas: while I think he is largely correct, I also think that the promise and pitfall of the postmodern worldview is that the reader is left to their own devices to construct their own meaning, which is as dangerously liberating as it sounds (but I shouldn’t have to tell you that). Azuma never really addresses this (although he was writing this in 2001 so he may have done so by now).

    • Also, so far as I can tell, that wikipedia link never actually defines what “iki” is. Can you help out there?

      • I have actually read The Structure of Iki (which I more or less linked to point out the similarities between the title and your title), and hopefully one of these days I will do a post summarizing it, more for myself than anyone else (I have the UK library copy sitting here; need to work on it like I need to work on everything else and ACTUALLY POST SOMETHING).

        The basic gist of iki is that it is (roughly) dandyism in the Edo-period merchant class, and could also be described as “coquettish flirting”; just think geisha and you’ll be close enough, I think. Kuki goes fairly in-depth into it, and says (and I am hugely paraphrasing here) that its “intensional features” (the personality assumptions that belie the observable iki behavior) are three things: bitai (coquetry), ikiji (pride, connotations of sophisticated), and akirame (resignation to fate, especially re: romantic relationships). It gets more complicated than I really want to get into here (and will save that for the summary post, that may yet come about).

        Of interest is that in Otaku Azuma spends a good twenty or thirty pages describing otaku as considering themselves the successors of the Edo merchant class and having constructed some kind of “virtual Japan” that resides in the Edo period. So the terms might be more connected than I thought originally. Even without that, though, I think looking at Kuki’s work could at least help clarify the issue for those of us who think entirely too much. :)

      • Oh, that’s really interesting. It sounds like the closest western analogue would be the decadents (think Oscar Wilde). I should try to ILL that book sometime. Or buy it. Whichever. : )

  2. Like everyone else, I really need to read that book. I know Pontifus found a copy in his library, but don’t know if he started it yet.

    I suppose I can see it as being a particularly postmodern conception, the reader building the reading from a “database.” I *feel* there’s a difference between such a building and the traditional interpretive model of reading (which would pre-date the postmodern period), but I’m not sure there actually *is* one. I haven’t examined the concept enough to do more than say what I feel, in my guts (where my guts are), so to speak.

    • The list price is $17.95, I would think even your perennially broke status could afford that.

      I feel there is a difference (although I don’t think the reader builds the reading from the database, the text is constructed from the database and the reader reads the text, and through the text generates meaning, if that makes sense: the database is described somewhat like your concept of genre, last time I talked to you about genre) and there may be one; I do know that societal meaning is considered to have been destroyed in the postmodern world (I think, maybe; again, it feels like that), and my perception is that the way to cope in this world is for each individual to construct their own meaning through whatever the feel most strongly about.

  3. To not feel horror is not to mis-read the text, but to not see the undermining elements is.

    Doesn’t execution factor in this? Similar to how comedy works? Poorly made horror films as well as poorly delivered jokes fail to elicit emotional responses precisely because of execution. (I’m sure exceptions exist… but interestingly, the ones I can think of are people who find the markers automatically dreadful/horrifying e.g. haunted houses, attics, silent children with knowing eyes, ghosts, etc).

    Does moe work this way?

    What I mean is that any cute girl who trips and falls while earnestly attempting to do something not inherently difficult (put a book back on a shelf) has a constant chance (as markers go) of eliciting moe?

    Does the best way to increase the chance, is to add glasses; or put her in a french maid uniform… mix and match other elements from the Database OGT mentions?

  4. > Certainly there are typically markers in the text on which moe is built, but those markers are not, in themselves, moe. We have coded them as such in the fan-space, the viewing gestalt.

    Isn’t that a chicken and egg argument? Now that moe “markers” are codified to some extent, creators embrace them and use them intentionally in textual productions, so it’s no longer phenomenological. And there is a good argument to be made that most of them where first introduced by creators (of bishoujo games in the late 90s) with the specific intent of eliciting feelings in consumers akin to what we term moe nowadays.

    The parallel with horror is interesting. And I guess that the markers by way of which me now recognize horror as a genre with its own recurring themes, tropes and conventions appeared progressively and with non-negligible input from fan circles.

  5. coburn

     /  6 February 2010

    Personally I’d encountered moe characters some time before reading about the word, but have never really felt that the few common sentiments those characters had elicited in me were part of some vital category I just didn’t have a name for. Which may well just be some kind of individual non-protective character trait of mine. I imagine that all of our first encounters with the label fit somewhere along the spectrum between the word ‘moe’ as a revelation and ‘moe’ as a seemingly arbitrary set of connections and well-embedded fetishes to learn. [maybe moe for Westerners could be compared to romantic love as a cultural export?]

    I’d be inclined to say that rather than moe being either in the fans or in the text, it’s either in the fans or (perhaps latently) in people in general. Because of my own experience I’m more inclined to see moe as a speciality fan fetish that exploits a relatively minor sentiment, as opposed to something undeniabe and primordial (like horror) – although (bearing in mind OGT’s comments) how far that distinction matters I’m not at all sure. In the same way that a transgressive horror film can tap a primordial fear surely a well of potential moe can be tapped, the fetishes of tomorrow gestating in pop culture today? A new form of moe might just need one good show to catapult it into the firmament.

    • ubiquitial

       /  12 February 2010

      Interesting. I will take this another step. There is something inherent about the psychology of Otaku that creates an innate affinity for what we know as “Moe”. Perhaps if we can understand why they possess this affinity, which is not as prevalent in other people, the nature of Moe will become apparent.

  6. @mt-i: I certainly think there’s a circle of influence happening between creators and fans, which complicates my claims — if it doesn’t short them out altogether. Here’s a kind of practical application philosophy: if the markers themselves ARE moe; that is, if they have an inherent, essential quality that IS moe, then you’re simply wrong if you don’t find a girl with glasses who trips a lot as moe. If you don’t feel a desire to protect her (the feeling in the audience is usually what moe is defined as), then you are reading the text incorrectly. If, on the other hand, the markers described are typically used as the locus of phenomenological moe construction, then if you do not feel the urge to protect, you can continue to interact with the text.

    It’s basically a specific application of the typical insistence that nothing is *essential*.

    Also, as Bacon would say, our terms are slippery and not working to describe anything. Can both the elements that elicit a feeling and the the feeling itself be moe? That’s how we use the word currently, and in practice it works just fine — but conflating the two does lead to a few problems, which I’m attempting to outline. Effectively, it makes more sense (I argue) to separate the two values, the locus site and the fan-feeling, because we can’t treat both in the same manner, one being firmly in-text and the other being actualized outside it.

    @coburn: I think I might read moe the way you do, or similarly — typically it’s an element in a show, for me, that’s already doing something else. And it’s usually not that strong, as I tend to prefer characters who can protect themselves, or are learning to (which makes it a bit of a paradox that I study the Gothic, of course, but then I’m mostly interested in the 19th century stuff [the wilting heroine is the main protagonist of 18th c. Gothic]).

    @ghostlightning: Execution must necessarily play a factor, but can we effectively analyze something that we are also arguing isn’t done well? I know the practical answer is yes, of course, but if we’re claiming something is flawed in the text, any attempt to reconcile the elements runs into the wall of “Well, it just wasn’t well-written [directed/acted/whatever].”

  7. Pontifus

     /  10 February 2010

    Well, I think every commenter so far has mentioned something I would’ve mentioned if I’d gotten there first, but I’ll try not to overlap too much.

    What it comes down to, I think, is, as Cuchlann says some few hundred pixels up, separating the values of structural moe and fan response moe — that is, discerning from the one word its two functional definitions, and I think it’s pretty clear that there are indeed two. We also need to scrutinize those definitions, to figure out how useful they are. Certainly “moe” has proven useful as a descriptor for the way one feels about a particular character, a personal kind of response — in this sense it happens in phenomenological space, which of course is shaped in part by the cultural experience of fandom, and, if I’m understanding and appropriating Azuma correctly (having read the essay precursor to Otaku and not the book itself), the experience of fandom, of fans throwing opinions at each other and “canonizing” the more consistent opinions, is where the database comes from.

    But the kind of moe I generally find myself most interested in is that other kind, structural moe, which, I have to admit, seems to some degree less tenable as a concept. Let’s say we’re given a character who appears to require protection of some kind, whether immediately or in future episodes/chapters/what have you. This character happens to be genuinely endearing enough that a fair number of consumers don’t hate her (to use the gendered pronoun applicable to most cases). Is she moe? Let’s say you do happen to dislike her. It no longer matters that she needs protection if you don’t want to be the one protecting her. If she isn’t moe to you, can you still make a broad claim as to her structural moe?

    In the interest of full disclosure: in that Sora no Woto post of mine that Cuchlann linked, I was really thinking more in terms of fan (one fan, i.e. me) response — though it was a response to a broader “segment” of structure than one character in isolation. Structural moe probably comes in when characters, as structural elements, interact with other elements to illicit the moe response — but I’d imagine the moe response is still phenomenological. I worry about Kanata’s fate on a battlefield; other viewers may not consider combat a possibility at all.

    Now, what I think is possible to judge, to some degree, is moe as it functions within the textual system. In terms of characters, who feels protective toward whom? Why? If character A feels the need to protect character B in that moe sort of way (whatever that means), and character A is sympathetic to population X, we might assume with some justification that population X might be somewhat likely to find character B moe (or feel moe for character B, as it were). But in this case I’m not sure we’re even talking about moe-as-such anymore.

    • ubiquitial

       /  12 February 2010

      I don’t think that the concept of “Moe” can be summed up as just “an urge to protect”. Nor do I think it should be referred to as such.

      • Pontifus

         /  12 February 2010

        It’s a specific kind of protectiveness, one that should be distinguished from protectiveness generally. But as I understand it, moe has indeed come to be identified by fans with a protective urge, as the feeling of an older sibling for a younger, or a mentor for a pupil, if only to distinguish it from sexual attraction, which moe is not. Of course, moe and sexual attraction can coexist — but, in my experience, it’s usually an uncomfortable alliance.

      • ubiquitial

         /  12 February 2010

        No sexual attraction? As far as I know, the intentional inclusion of “Moe” in Eroges is one of the driving forces behind the industry. In fact, Sexual attraction is a prevalent, if not essential, aspect of “Moe”.

        And as for “the urge to protect”, I think that can also be translated as the “urge to violate” (rule 43). I mean, with all the tentacle rape hentai out there…

      • Pontifus

         /  13 February 2010

        I love that Cuchlann used Mikuru Asahina as his eyecatcher, as she’s an excellent example of moe and attraction thrown together into an uncomfortable zero-point. Consider Kyon’s attitude toward her. He’s obviously attracted to her, and he demonstrably feels the need to protect her because she’s cute and helpless — but he tends not to experience both at once. When he’s titillated, he more or less lets Haruhi have her way with Mikuru, and contents himself with apologies he never actually voices; this is how most of the costume incidents go, if I remember correctly. But when he does defend her from Haruhi’s machinations, sex never seems to be the closest thing to mind — though, as often as not, moe and attraction battle it out in his mind, at which times Kyon tends to feel at least a little guilty. And he’s always worried that his attraction to Mikuru will put him at odds with the school’s moe-driven male population, which, for the most part, hovers protectively around Mikuru, but tends not to move in with sexual advances.

        As far as eroges go, I tend to think that moe is not operative during sex scenes. Moe is definitely a significant factor at play in ren-ai games, but it isn’t the be-all, end-all driving force behind characters.

      • ubiquitial

         /  13 February 2010

        But again, can you really simplify “moe” as the urge to protect?
        Then how do you explain the Moe characteristics: Zettai Ryouiki, Megane, Tsundere, Twintails, Etc. That have nothing to do with protection?

        And I disagree with you in that I believe “moe” and attraction can coexist.

        In the end, it’s how you define “moe” that determines it’s nature. But “moe” is just such a hazy concept. Lets go back to the beginning, shall we. Forget everything we’ve come up with about “moe”.

        Name a characteristic of something that is “moe”.

      • Pontifus

         /  13 February 2010

        I wouldn’t identify zettai ryouiki as moe (unless the point of it is the exposure), and, rather than a moe characteristic, I’d call twintails a tsundere indicator. But megane and tsundere have something in common: the character to whom they apply is flawed. A glasses-wearing character can’t see without external aid; the glasses are a kind of self-protection — and besides, I’ve seen glasses used too often in anime and manga as a metaphoric protective barrier between the meganekko and the “real” world to discount that interpretation as well. And a tsundere character requires protection from herself, usually from her own lack of confidence and her inability to form meaningful romantic relationships (or, more rarely, meaningful relationships altogether).

        Sure, moe and attraction can coexist, but I maintain that they’re often in conflict with one another when they do.

        You’re absolutely right in that moe is as ill-defined as pretty much any abstract fictional concept, and that we’re all going to have different definitions at the end of the day. With so many people using the term, there’s really no avoiding that. I suppose I wouldn’t simplify moe into “the urge to protect,” which is too broad — but I would simplify it by saying that moe is brought about by a degree of helplessness in a character (hence that crowd that denounces moe as irreconcilably sexist), so protectiveness is certainly a possible factor, and, as far as I know, a pretty common one.

      • ubiquitial

         /  13 February 2010

        But again, all of this is futile if we cannot define “moe”. There’s no point in figuring out its relationship to attraction if we cannot understand “moe” in itself.

      • Pontifus

         /  13 February 2010

        Well, I think there are two ways we might go about “defining” moe. Firstly, we could come up with a functional, temporary definition based on how a number of people seem to use the word, with the understanding that our definition would be specific to whichever population we’re mining. And secondly (and this may be the same as the first, essentially), we can come up with a theoretical definition, test it against the common use of the word, hang on to it as long as it’s productive in terms of fueling thought, and throw it out when we don’t need it anymore. So, even if moe is technically undefinable, I don’t think trying to define it is necessarily a pointless endeavor.

      • ubiquitial

         /  14 February 2010

        So, as long as something has a function, it can be defined in terms of words? Define Love. Define Desire. Define Beauty. Define Consciousness. Define Rationality.

        As I see it, “moe” is more of an emotional response, and therefore cannot be put into words. Since emotion and rationality belong to irreconcilable sectors. Put simply, it’s that feeling we have when we see a “moe” character, and nothing else.

      • coburn

         /  14 February 2010

        Presumably what separates moe from those other indefinable things is that it can seemingly be defined, if not by words then by signifiers/tropes. It is and it isn’t subjective. If I follow Pontifus right and he reads moe in a given character as inevitably rooted in protectiveness then it would suggest some kind of conditioning process through which that urge is buried within those signifiers.

        If it is buried thoroughly enough then we can potentially find moe in something without feeling it properly – presuming, as ghostlightning was suggesting, that it is possible for the signifiers to be delivered sufficiently badly.

        I mean, glasses wearers aren’t really helpless (I hope)? If glasses cause moe then either they are subliminally moe or we have all just learnt to be experts in wrong. So the ‘structural moe’ argument comes down to an identification of helplessness as a root cause rather than a direct urge. i.e. the helplessness of the character is potentially occurring on a separate psychological level to the conscious moment where we feel moe. Protectiveness toward the capable person becomes possible. So maybe ‘structural moe’ = hidden protection fetish, and ‘fan response’ = customary anime fetish. I’m not sure which of those is worse, although I don’t think that such a deeply rooted protectiveness is necessarily misogynistic.

      • ubiquitial

         /  14 February 2010

        “That it can be seemingly be defined” is still a very ambiguous statement.

        And I agree with you in that there is some level of conditioning involved, but that seems altogether unlikely, since that would require the same characteristics to be repeated over and over throughout cultural media of all forms, by coincidence. I believe it’s more of a collective unconscious that governs what we see to be “moe”

      • Pontifus

         /  15 February 2010


        Define Love. Define Desire. Define Beauty. Define Consciousness. Define Rationality.

        1. Population x seems to employ [love/desire/beauty/consciousness/rationality] with functional definition y; definition y is useful when we’re talking about population x.

        2. Let’s define [love/desire/beauty/consciousness/rationality] as z, and test z for its usefulness.

        This is what I’m suggesting we do with moe. Without this kind of inquiry, there’d be no philosophy, humanities, or social science.


        Well, at the end of the day, I’m really with Cuchlann in that moe is “located” in the fandom, rather than the art — but at the same time, it’s possible to point out signifiers that will probably trigger moe in some people, and a major commonality seems to be helplessness/the need to be protected. Not that helplessness and moe are so inseparable that other factors can’t intervene and preclude the moe response, though.

        So the ’structural moe’ argument comes down to an identification of helplessness as a root cause rather than a direct urge. i.e. the helplessness of the character is potentially occurring on a separate psychological level to the conscious moment where we feel moe. Protectiveness toward the capable person becomes possible.

        On a separate psychological level — yeah, I think that’s it. Fans bring moe; a signifier becomes a moe indicator when enough fans use it to signify moe. So probably structural moe and fan-response moe are the same thing insofar as structural elements become moe through accumulated fan activity. It may be more useful to define moe in terms of external (fan feels moe for character) and internal (character feels moe for character) — but in the latter case we need a functional definition of moe, and that always seems so hard to come up with.

        @ubiquitial again

        I believe it’s more of a collective unconscious that governs what we see to be “moe”

        It’d be difficult (to say the least) to identify where moe comes from ultimately. But I imagine it has to do with the set of psychological and social factors that lead people to the fandom in the first place. Personally I bet it’s related to the otaku fascination with “pure” women/virgins — but I’m not ready to get into that yet.

      • ubiquitial

         /  15 February 2010

        Pragmatic Bells are Ringing~ Pragmatic Bells are Ringing~

        Ding Dang Dong~! Ding Dang Dong~!

      • ubiquitial

         /  13 February 2010

        Also, I wouldn’t attribute Kyon’s awkwardness with Mikuru as a conflict between moe and attraction, but just awkwardness in general.

      • Pontifus

         /  13 February 2010

        Ah, but as someone who is in “real” life almost absurdly awkward, I know well that awkwardness has to come from somewhere, and, for me, that “somewhere” tends to be some inner conflict. I’m not saying that inner conflict is the source of awkwardness, mind you, but it’s definitely a source. And, sure, maybe the source of Kyon’s awkwardness is simply a lack of experience with women, he being a teenager fresh out of middle school and all. But the problem of reconciling moe with attraction could, as I see it, arise from a lack of experience with women, as well; a more experienced guy would already know to which degree each is viable (i.e. he’d know that moe is rather idealistic).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 372 other followers

%d bloggers like this: