Anonymous asks:

Thoughts on the Pokemon Tree of Life? (Search on Reddit.) I just spent a good half hour marveling at how much work must have gone into it!

I’ve seen an earlier version, I think in generation V or VI. And, well, I can certainly appreciate the effort, and it is a beautiful thing in its own way. Sort of artistic, really. The trouble I have is that it almost certainly doesn’t work like that. Two kinds of animal that can breed to produce viable offspring (e.g. horse and donkey) almost have to be more closely related to one another than two that can’t (e.g. horse and goat). But here we get Pokémon that can breed with each other despite being from distant parts of the tree, like Tangrowth and Venusaur, and Pokémon that can’t breed with each other despite being adjacent on the tree, like Gardevoir and Medicham. Pokémon clearly don’t follow the real world’s rules of genetics, heredity and speciation; they can breed across species lines and they can inherit at least some acquired characteristics (Lamarckian evolutionists represent!). What that means is that, if it’s even possible to come up with a schema that represents the evolutionary relationships of all Pokémon in a consistent and meaningful way, it probably looks nothing like a real-world “tree of life” or cladogram. It’s… probably more of a Venn diagram situation. Or, like, a spirograph or something.


How do you think Pokemon breed?

…hoooooo boy.

So I wrote this thing about it once upon a time, and I think the big takeaway from it should be not so much anything I actually said, but that we know so little, and what we do know makes such little sense, that you can pretty much say just about anything and have it be on some level justifiable.  Game Freak has always been quite insistent in telling us that there’s an awful lot about Pokémon reproduction that no-one knows at all.  Back in the original Gold and Silver, a Pokémon egg was supposedly a major discovery because before then, no-one actually knew for sure that Pokémon hatched from eggs, which sort of makes you wonder how Pokémon ranchers and the like ever managed to do their jobs.  Then in X and Y, sensing that we foolishly thought we knew what we were talking about, Game Freak decided to remind us that we don’t know a damn thing by telling us that Pokémon eggs aren’t really eggs.  At this point you could tell me that Pokémon come from spontaneous generation and I wouldn’t quite be willing to rule it out (I mean, I’m also on the record as suggesting – only half-jokingly – that Pokémon originally evolved from rocks, so it’s not like I’ve got a whole lot of credibility to protect here).

VikingBoyBilly asks:

In the episode Extreme Pokémon, the day care man gave ash a (teal? blue?) egg in a glass case and he said “when the pokémon hatches, use the pokéball on top of the case to hatch it with.” So… is that what’s happening when you receive eggs that already have a pokéball from the day care man? (incidentally, was that the larvitar egg, or is it another pokémon?)

(I don’t know the dialogue from that episode offhand, but I think you mean to say “use the Pokéball to catch it with,” not “hatch it with,” because if Pokémon actually cannot hatch without a Pokéball then we have some serious problems here)

I suppose it must work something like that?  I mean, we can hatch eggs even if we have no Pokéballs in our inventory, and the baby Pokémon have Pokéballs automatically, so unless we envision Pokémon somehow hatching with Pokéballs, someone must be supplying free ones with every egg.  You can probably read into this, if you choose, all kinds of sinister things about being born into slavery (which could certainly be a very interesting way to take it), but I don’t think you have to for it to make sense.  If you think of the main functions of Pokéballs being protection and transport… well, no one wants the most vulnerable Pokémon on their team to be forced to walk everywhere and have nowhere to retreat to in case of danger or injury.  And the alternative – just releasing an infant Pokémon into a potentially hostile environment with no caregiver because you happened not to have any Pokéballs at the time – is clearly lunacy.  I mean, in practice we do that in the games all the time and in astonishing numbers, but you sort of have to give them points for trying…

(also I believe the egg you’re referring to is the one that eventually hatches into Ash’s Phanpy)

vikingboybilly asks:

Okay, listen to this: What if, instead of breeding eggs, pokemon trained each other by playing at the day care, giving their IVs, egg moves, or abilities to the other pokemon, and your pokemon can reach it’s full potential by learning and growing from the other pokemon at the day care instead of ridiculous egg breeding chains? This could be controlled with the power items and everstones and such, but it might work a little differenty (The power bracer’d pokemon get the attack IV from the other

You know, I think I like that, or something along those lines anyway.  I talked a bit in one of these old things about expanding the complexity and interactivity of Pokémon storage, placing a bit more emphasis on what your Pokémon are getting up to when they’re not actually with you, and this kind of thing might play into that kind of aim really well.  I think it reinforces the sense you get of Pokémon as individuals if they help each other to learn and grow, rather than just keeping it as the sole responsibility of the trainer.  There’s a lot of scope there for, say, the combinations of Pokémon’s natures having different effects on each other, or for abilities that (in addition to their battle effects) alter the way Pokémon can interact with each other in the day care.  Maybe Pokémon that have trained together will be more effective in double or triple battles together?  Lots of different directions you could take it.

Pokémon and Gender

So, I’ve been wanting to write this entry for a while, but haven’t because I can’t make it fit into any of the series I want to do.  In that sense, it’s actually something of a cool opportunity that I’m not committing to writing anything in particular at the moment, because I can just do whatever.  I will warn you, though, that this will be one of my most trippy and speculative entries yet.  Brace yourselves.

The premise of what I’m going to be talking about today is a choice of vocabulary that just about every person on the planet has probably taken for granted, but which has always stuck in my craw (because, as we know, I’m obsessed with languages): “gender.”  The word “gender,” unbeknownst to many, doesn’t actually refer to the biological distinction between male and female.  Male and female are sexes, not genders.  Masculine and feminine are genders.  Traditionally the word has referred to the concept of grammatical gender, an idea present in every major European language except English, whereby all nouns are considered masculine, feminine, or (in e.g. Latin, Greek, and German) neuter, and adjectives must change their forms to suit the gender of the nouns they describe (for instance, in Latin, ‘tall’ is altus if you’re talking about a mountain, but alta if you’re talking about a tree).  There is no real rule to these, and they’re not always consistent across languages either (the Latin word for tree, arbor, is feminine; the ancient Greek word, δένδρον, is neuter, and the French word, arbre, is masculine, even though it’s clearly derived from the Latin word).  In short, something that has ‘gender’ is associated with maleness or femaleness (or absence thereof) in some vague and unspecified way.  In modern usage, this meaning has expanded to include descriptions of a person’s general psychological disposition, certain traits being regarded as ‘masculine’ – typical of a man, but not exclusive or necessary to men – others as ‘feminine’ – typical of a woman, but not exclusive or necessary to women.  We’ve all met masculine women and feminine men; I’ve been called ‘feminine’ once or twice (by my best friend, no less).  Things get much worse when we throw sexuality into the mix, because that’s even more complicated and doesn’t necessarily line up with sex or gender, but honestly I don’t want to go anywhere near that particular can of worms.  Anyway, here’s the thing…

Pokémon don’t have sexes.  They have genders.

I’m aware, of course, that this is probably a mistranslation caused by squeamishness about exposing ten-year-olds to the inherent horror and immorality of the word ‘sex’ (never mind that you’re selling them a game in which Pokémon engage in ‘breeding’).  I think we can all agree, though, that just letting it go would be far less entertaining than grabbing it with both hands and following it to its most insane possible conclusion.

The Pokémon franchise in general has always been extremely closed-mouthed about how Pokémon reproduce.  None of the characters seem to have any idea how the process works.  A number of NPCs pointedly insist that it hasn’t been proven that Pokémon lay eggs, because in thousands of years of recorded history no-one has ever actually seen it happen.  Until the events of Gold and Silver, likewise, it hasn’t been conclusively proven that Pokémon hatch from eggs either; that’s why Professor Elm is so excited when your Togepi egg hatches.  A Pokémon egg is an incredibly rare curiosity, the preserve of obsessive collectors like Mr. Pokémon.  The kind men and women of the various day-care centres, likewise, are utterly mystified whenever eggs show up in their backyards, no matter how many times it happens.  There is also a whole string of little discrepancies in the system as it’s given to us.  There are a few single-gendered species, which creates obvious problems – female-only Pokémon, like Kangaskhan and Lilligant, would need to breed with males of other species in the same egg group to maintain a population (because, of course, inter-species breeding is not especially problematic for Pokémon), while male-only Pokémon can’t even do that, since all baby Pokémon are of the same species as their mothers; players can only get babies of those species with the help of the ‘breeder’s wildcard,’ Ditto.  According to everything we have been told, Sawk, Throh, the Hitmon triplets, Braviary and possibly Tauros cannot reproduce in the wild.  The same goes for ‘genderless’ Pokémon, like Electrode and Starmie.  Now, I realise biology has never exactly been Pokémon’s strong point, but it doesn’t take a genius to work out that a species which is completely incapable of reproducing cannot exist.  It doesn’t seem to make a whole lot of sense for Chansey, Petilil, Mandibuzz, Kangaskhan and Jynx to be totally reliant on hybridisation to continue their species either.  There’s a little question mark over Ditto as well – Ditto can Transform into an exact duplicate of a Pokémon standing in front of it… so, by all common sense, a Ditto presented with a biologically male Pokémon should Transform into a biologically male Pokémon.  Ditto might be able to alter its form to a small extent, but the games don’t really provide any evidence for that, and the anime implies that it can only make superficial changes, so I doubt it could reconfigure an entire organ system without help.  In short, whatever goes on in day-care centres, it’s not straightforward sexual reproduction on the model of real-world animals.

Here’s my weird-ass take on it all, then…

Pokémon, I will repeat, don’t have sexes.  They have genders.  That is, they don’t actually have differentiated reproductive systems; they are all, in essence, single-sex species.  They do have an unusually large degree of variance in the levels of different hormones they produce, which leads to significant variation in their psychological traits, and in many species (most notably Nidoran) this is linked to some physical aspects, creating the appearance of sexual dimorphism, though in the vast majority of cases the differences are actually superficial.  Reproduction takes place via a ‘mind-meld’-like process (I sort of imagine them pressing their foreheads together, murmuring to each other, and glowing softly); genetic information is exchanged, but selectively – the vast majority of a baby Pokémon’s genes come from its mother.  Most of the exchange actually involves psychological traits.  As a result, a baby Pokémon will be quite close to being, physically, a clone of its mother (which is why inter-species breeding always results in a Pokémon of the mother’s species – the father contributes only a few genes, selected out of those that are compatible with the mother’s species) but will have closer to an even mixture of psychological traits from both parents.  The father (as, of course, we know) is additionally capable of passing on a number of conscious mental traits and learned abilities, which become ingrained in the child’s instincts.  For most Pokémon species, mental health requires a mixture of masculine and feminine traits, so instinct dictates that two masculine Pokémon will not mate willingly, and nor will two feminine Pokémon.  The entire process is far more low-key than what real animals have to go through, and consequently much more difficult to observe, which is why the whole subject is surrounded by such abject confusion.

So, how does this help to resolve the problems with how Pokémon breeding appears to work in the games?

The thing about the all-masculine species, like Hitmonchan and Braviary, is that – being universally and excessively ‘masculine’ – they are extremely pugnacious and aggressive (this, again, is something we already know – just look at the all-masculine species).  As a result, practically everything they do is constantly simmering with potential to break into outright violence.  What passes for ‘courtship’ among these species is no exception, and is simply so confrontational that human observers have never actually made the leap to identifying it as courtship (if you’re familiar with Homestuck, the concept of a ‘caliginous romance’ is a decent analogy, though it’s far more developed and laden with cultural baggage) and, as I suggested, the actual reproductive act itself is surprisingly easy to miss.  The kind of aggression and conflict necessary for a pair of Pokémon from an all-masculine species to develop an intimate relationship simply isn’t allowed to happen in the context of a day-care centre, where the staff normally discourage fighting.  Thus, Pokémon from all-masculine species can and do reproduce in the wild, but never get the chance in a day-care.  Pokémon from all-feminine species have a similar, but opposite set of issues.  They are universally and excessively ‘feminine,’ and therefore extremely passive, gentle, and cautious in their relationships with each other.  Courtship is an extremely slow, drawn-out process that can last for months or years; in captivity, there normally just isn’t time to observe it happening, and even in the wild it’s so long-term that human scientists haven’t actually been able to recognise it yet.  Many Pokémon from all-feminine species will take masculine partners from other species in order to create social diversity, and this generally happens much more quickly.

‘Genderless’ Pokémon are another kettle of fish entirely.  I want to suggest that they don’t necessarily all work in the same way; rather, they’re a ‘miscellaneous’ group.  Many of them aren’t actually ‘genderless’ but actually have three, four, five or even more genders, none of which match up exactly with ‘masculine’ or ‘feminine’ – as a result, humans are totally unable to understand the rules that govern their reproductive compatibility.  Some of them reproduce in groups of three or more, making it impossible for a day-care centre, which takes two Pokémon at most, to observe their reproduction.  A few reproduce by fission, splitting into two or more children only at the moment of death.  In short, their reproductive practices are just so weird that human observers don’t have a hope in hell of understanding what’s going on.  This, of course, brings me to the most important Pokémon of the lot: Ditto.  Ditto, in the games, does not reproduce; it only helps other Pokémon to do so, presumably using its ability to Transform into any other species.  Although Ditto forms a perfect physical copy of its partner, psychologically it doesn’t change at all when it Transforms; since Ditto is neither masculine nor feminine, it adopts a totally different role and all the usual rules of courtship go out the window when it gets involved, which is why universally masculine Pokémon can reproduce in captivity with a Ditto.  Ditto is likewise capable of overriding whatever whacked-out reproductive norms are in play for ‘genderless’ Pokémon, even producing eggs of Pokémon species that don’t naturally lay eggs at all.  It contributes very little to the child, physically or psychologically, but does provide a way to scramble the genes provided by the other parent and throw up new combinations of dormant traits.  So, then… question: why, in evolutionary terms, does it make sense for a species to focus its energy on helping other species to propagate themselves?  Answer: the relationship is symbiotic.  Ditto actually feeds on the leftover energy of cell division to revitalise its own cells.  It gets… a bit metaphysical, but the practical result is that, as long as a Ditto continues to help other Pokémon reproduce, it will never die.  Absorbing a huge excess of cellular energy allows Ditto to split and form two new Ditto; this doesn’t happen often, but accounts for the Ditto that are inevitably killed by other Pokémon or die in accidents.

As for where the Ditto came from in the first place, I’m inclined to accept the fan theory that they’re closely related to Mew – the only other Pokémon with the ability to Transform, courtesy of her genetic library, who also happens to be bright pink – mostly because it fits well with my ideas about Mew, which suggest that her whole purpose in the world is to absorb DNA from other Pokémon and store it.  Ditto have lost the ability to store borrowed DNA on a long-term basis, and as a result their physical form has degenerated, but they retain the ability to absorb DNA and rapidly assimilate it into their own systems.  That, naturally, brings me to the last category of Pokémon I need to talk about: legendary Pokémon, who (with the notable exception of Manaphy) cannot breed at all, Ditto or no Ditto.  Most of them are also genderless.  Many legendary Pokémon are heavily implied to be unique (and presumably immortal) anyway, which means I don’t have to worry about them, but a few seem to exist as entire species; most significantly, a baby Lugia appears in a few episodes of the anime.  They’re sufficiently different from other Pokémon that they can’t breed normally with anything else, and their lifespans are so long that humans just can’t observe them properly.  Mating season might come around once every three or four centuries and last for a month or two; even then, eggs might take years to hatch.  In short, some legendary Pokémon do breed, but for all intents and purposes it’s not something humans can take advantage of.

I think that’s enough from me for now – it’s not every day I try to totally redefine the way we look at a major aspect of the Pokémon games.  As always, I’d love to hear your thoughts on my latest weird-ass theory; if you point out something that doesn’t make sense, I might be able to improve on it.  Anyway, that’s all from me – thanks for reading, and have a fun day!