Post History

What does "parameterised type in the positive / negative position" mean in the context of invariant functors?

From PureScript's [`Data.Functor.Invariant`][1] documentation (emphasis mine):

> A type of functor that can be used to adapt the type of a wrapped function where **the parameterised type occurs in both the *positive* and *negative* position**, for example, `F (a -> a)`.

What do the terms "_positive_" and "_negative position_" refer to above?  
<sup>.. and also wondering why `F (a -> a)` requires an entirely new type of functor, but that should be another question probably..</sup>

I tried to figure this out both by learning more about invariant functors and looking up the terms "positive", "negative" in connection with functional programming, but wasn't able to connect the dots yet. Is there a basic resource that I've missed? Or, at least, where should one start to understand these concepts? (Not even sure what the right topic is; see my research attempts below.)

---

From the search results for 
  ["invariant functor positive negative"](https://www.google.com/search?q=invariant+functor+positive+negative&oq=invariant+functor+positive+negative&gs_lcrp=EgZjaHJvbWUyCQgAEEUYORigATIHCAEQIRigATIHCAIQIRigATIHCAMQIRigAdIBCDU4NDFqMGo3qAIAsAIA&sourceid=chrome&ie=UTF-8)
, ["positively invariant vs negatively invariant polarity"](https://www.google.com/search?q=positively+invariant+vs+negatively+invariant+polarity&oq=positively+invariant+vs+negatively+invariant+polarity&gs_lcrp=EgZjaHJvbWUyBggAEEUYOTIHCAEQIRigATIHCAIQIRigATIHCAMQIRigATIHCAQQIRigAdIBCTEzNDMzajBqNKgCALACAQ&sourceid=chrome&ie=UTF-8)
, ["type theory positive negative polarity invariant"](https://www.google.com/search?q=type+theory+positive+negative+polarity+invariant&oq=type+theory+positive+negative+polarity+invariant&gs_lcrp=EgZjaHJvbWUyCQgAEEUYORigAdIBCTExMTUyajBqN6gCALACAA&sourceid=chrome&ie=UTF-8)
, and ["what are invariant functors"](https://www.google.com/search?q=what+are+invariant+functors&oq=what+are+invariant+functors&gs_lcrp=EgZjaHJvbWUyBggAEEUYOTIGCAEQRRhA0gEINDM0OWowajeoAgCwAgA):

+ [`nLab`] [polarity in type theory](https://ncatlab.org/nlab/show/polarity+in+type+theory)

  My intuition tells me this is the answer - but I have too little math / logic background to understand most of it (and to see how the parts that I do get connect to my question).

+ [`blog`] [Polarity in Type Theory](https://existentialtype.wordpress.com/2012/08/25/polarity-in-type-theory/)

  Same.

+ [`stackoverflow`] [Example of Invariant Functor?](https://stackoverflow.com/questions/22103445/example-of-invariant-functor)

  The answer does mention the words "positive" and "negative", but it doesn't explain them, and neither do the linked resources. It does mention "HOAS",

+ [`google search`] [higher order abstract syntax HOAS](https://www.google.com/search?q=higher+order+abstract+syntax+(HOAS)&sourceid=chrome&ie=UTF-8)

  but it doesn't look like what I'm looking for.


+ [`hackage`] [`layers` package documentation 
 -> invariant functors](https://hackage.haskell.org/package/layers-0.1/docs/Documentation-Layers-Overview.html#g:14)

  Even if it is relevant, I don't get it...

+ [`ploeh blog`] [Invariant functors](https://blog.ploeh.dk/2018/03/19/functors-applicatives-and-friends/) and [Functors as invariant functors](https://blog.ploeh.dk/2022/08/01/invariant-functors/)

  No mention of "_positive_" or "_negative_", but these posts did lead me to Mark Seemann's articles series [Functors, applicatives, and friends](https://blog.ploeh.dk/2018/03/19/functors-applicatives-and-friends/) that looks compelling and comprehensive, but so far I didn't find anything there either (or I missed it..)

+ [`nLab`] [invariant](https://ncatlab.org/nlab/show/invariant)

  Um, way above my head...


  [1]: https://pursuit.purescript.org/packages/purescript-invariant/6.0.0/docs/Data.Functor.Invariant