The Canon Is Porous

A Gateway Philosophy for Global Cinema

A canon is useful until it begins to resemble a closed room.

I have already made the case for why a film like Hard Boiled belongs in the Great Cinema Master List.

That answered the question of admission.

What I have not answered is where a film is supposed to take you once you have seen it.

Selection asks whether a film has earned a place. Porosity asks what that place is for.

A film can matter for what it contains. It can also matter for what becomes visible afterward.

That second kind of value interests me here.

Most canons are built like monuments. You arrive, admire, record the visit, and move on. Accumulate enough of them and film history begins to resemble an itinerary somebody else prepared for you.

I want something more restless.

The best recommendation does not merely satisfy. It creates another question. A film reorganizes your curiosity. A filmmaker you had never followed begins to acquire a shape. A country stops being a category on a streaming menu. A genre you thought you understood starts showing you where it came from.

Sometimes the effect is simpler. You finish one movie and suddenly want another.

That is enough.

Belonging is not the finish line. It is the starting point.

What Becomes Visible

Take Indian cinema.

Sholay and RRR occupy different historical moments, but both show how comfortably popular filmmaking can carry action, melodrama, music and spectacle together without treating them as competing impulses.

Neither film explains Indian cinema. That would be an absurd burden to place on any movie.

Their usefulness is almost the opposite. They make the insufficiency of a single encounter obvious.

After Sholay, hybridity begins to look less like a collision of genres than a native fluency. After RRR, what initially seems excessive can start to look like confidence: a cinema that assumes its audience can move quickly between emotional and stylistic registers.

Then Lagaan, Andhadhun, Tumbbad, and Gangs of Wasseypur stop appearing as disconnected recommendations.

They become different answers to a larger question: how much can popular storytelling contain before the category we have given it becomes too small?

One film cannot represent that terrain. It can make the terrain visible.

Korean cinema opens another route.

Burning turns uncertainty into atmosphere. Mystery becomes tangled with class, desire and the unstable things people imagine about one another.

Decision to Leave starts with another recognizable structure, the investigation, but allows romantic obsession to reorganize it from within.

Put the films near each other and the useful conclusion is not that Korean cinema possesses one identifiable style. The distance between them is more interesting.

Genre and art film stop looking like opposite territories. Suspense can carry social observation. Romance can inhabit procedure.

Chinese-language cinema produces another kind of opening.

Infernal Affairs is tightly wound around divided identity and psychological pressure.

Crouching Tiger, Hidden Dragon moves through an entirely different register, where landscape, restraint, desire and martial movement carry much of the emotional weight.

The films barely resemble one another on the surface, which is why putting them near each other matters.

A single canonical representative compresses a tradition. Two sufficiently different films begin to break that compression.

They remind you how much difference can sit inside a label such as “Chinese-language cinema.”

A gateway should not define a territory. It should make the territory harder to underestimate.

One Filmmaker, Many Paths

Sometimes the gateway is one filmmaker.

Satoshi Kon may be the clearest example on this list.

Perfect Blue, Millennium Actress, and Paprika are not merely three animated films connected by recurring themes. Together they reveal editing itself as a way of thinking.

In Perfect Blue, identity fractures across performance and perception.

Millennium Actress lets a remembered life move through film roles and historical periods without insisting that one layer remain neatly separated from another.

Paprika pushes the instability further, allowing dreams and waking life to infect one another until the boundary barely holds.

A gesture can begin in one reality and finish in another. A performance can become a memory, then become cinema again.

The cut is not merely connecting shots. It is thinking.

Watch Perfect Blue alone and you can admire it as a psychological thriller. Put Millennium Actress beside it and something else becomes visible: Kon is not simply switching genres. He is finding different uses for the same freedom of transition.

Add Paprika and the pattern becomes difficult to miss.

The discovery travels beyond Kon.

You start noticing how much ordinary continuity editing does to stabilize reality for us, how carefully most movies announce the border between what is remembered and what is happening.

Kon keeps removing those railings.

Afterward, other films can look different because one filmmaker has made you conscious of an assumption you did not know you were making.

A gateway is not homework or a prescribed route. It changes what you are prepared to notice next.

A List Moves Sideways Too

Rankings encourage vertical thinking: higher, lower, better, worse.

GCML cannot escape that. It is a ranked list. I make those judgments, and elsewhere in this collection I argue for why making them can still be useful.

But cinema also moves sideways.

A viewer sees one film and develops a taste for a form. A style suddenly has a history. An unfamiliar industry acquires names. Animation stops being an occasional category and becomes an artistic tradition with its own arguments and reinventions.

A number cannot express that movement.

It can tell you where I put a film. It cannot tell you where the film might send you next.

That is why these first two essays operate differently.

Beyond the Big Three goes deep. It asks one film at a time to defend its admission.

This essay moves outward.

Depth makes a case. Breadth opens a path.

What This List Favors

Every list leans. This one does too, and I would rather say that plainly than disguise preference as inevitability.

GCML responds strongly to cinema in which popular form and formal ambition coexist. It takes genre seriously when genre becomes a vehicle for invention. It treats animation as cinema without qualification. It gives substantial attention to international popular traditions that have often remained peripheral in English-language discussions despite enormous artistic histories and audiences.

Those are choices, not neutral outcomes.

Another curator could build an equally serious list with a different center of gravity: more experimental cinema, more documentary, more silent work, or deeper attention to regional traditions GCML maps only lightly.

That curator would not necessarily be correcting me. They might simply be drawing another map.

Curatorial space is finite. When equally worthwhile routes compete for room, this list sometimes favors the route for which fewer maps already exist.

That helps explain why it spends so much time with international popular cinema, animation, genre filmmaking, and newer work whose reputations have not yet had decades to harden. These are not automatically more deserving than well-charted traditions. They are places where another route may be more useful.

I do not think a personal canon has to duplicate every road already marked clearly elsewhere.

A finite list produces blind spots. An absence may reflect limited knowledge. It may expose a preference worth reconsidering. Sometimes a film simply has not forced the question strongly enough yet.

The films missing from this list are not necessarily absent because they are unworthy.

Some are already well mapped elsewhere; others belong to routes this project has not emphasized. Inevitably, a few have simply been overlooked, and others may eventually demand a place as the list evolves.

Every map has a maker. That does not grant immunity from criticism. If anything, it increases the responsibility to make the choices visible.

The Working City

There is a familiar fantasy behind canon-making.

Enough research, enough viewing, enough corrections and eventually the map will be finished.

It will not.

Cinema is too large and too dependent on survival, restoration, distribution and luck.

Films disappear. Others return. A print is restored. A director receives a retrospective. A movie barely available in one decade becomes easy to see in another. New films arrive and complicate categories that had begun to feel settled.

Reception moves too. A film dismissed on release may acquire weight decades later. A style that once seemed isolated acquires descendants, and suddenly the older film looks different because the films around it have changed.

The canon is not a completed building.

It is a working city.

Some streets are crowded. Others are poorly marked. Some close for years and reopen. A route that means nothing to one viewer can become essential to another.

The map matters, but it cannot substitute for moving through the place.

That is why I do not measure this part of GCML by whether anyone reproduces my choices.

I measure it by movement.

Does Sholay make someone curious about a film culture larger than the small set of titles they already knew?

Does Burning change what they expect a thriller to do?

Does Perfect Blue lead to Millennium Actress, then Paprika, and eventually to a different awareness of editing in films that have nothing to do with Satoshi Kon?

Maybe. Maybe not.

A viewer may follow an entirely different thread. They may watch Sholay and become fascinated by Amitabh Bachchan rather than by genre hybridity. They may leave Burning wanting more Lee Chang-dong. They may see Paprika and head toward experimental live-action cinema instead.

Good.

A useful map does not decide the journey.

Keep Looking

There are two bad fantasies available to anyone building a canon.

One is that judgment is impossible, so selection should be abandoned. The other is that judgment can become complete, so selection can finally end.

I believe neither.

We have to choose while remaining capable of discovering that our choices were incomplete.

That tension is not a defect in GCML. It may be the only way the project remains alive.

I want the list to contain convictions strong enough to argue with, but I also want its edges to remain visible.

A viewer should be able to follow a film beyond the rankings, beyond the titles collected here, even beyond the assumptions that led me to include it in the first place.

GCML is one map.

It should point toward roads beyond itself.

Its purpose is not to settle cinema history. It is to give curiosity somewhere to go.

Eventually, if the map has done its job, the viewer no longer needs it.

The aim is not to complete the canon.

It is to keep it open.