Titanic (1997) Review
8/10
I’ve watched this film more times than I would have predicted in 1997, when I assumed it was a phenomenon I’d be tired of within a year. It’s outlasted that assumption and most of the objections raised against it, including several of mine.
The premise is the simplest one available. Two people from different classes meet on a ship, have three days, and the ship goes down. Everyone in the audience knows the ending before the lights dim. That’s not a weakness Cameron worked around. It’s the engine.
The 8 accounts for what holds and what doesn’t. The technical achievement is genuine and enormous. The class material is broad to the point of caricature, the first-class passengers rendered as cartoon villains while steerage dances. The door is a real flaw and I’m not going to pretend otherwise. Against all that, the film makes an ending everybody already knows land anyway. That’s the hardest thing in narrative, and it does it for three hours without the tension slackening.
The Ship Is the Argument
The Titanic was built as a statement about what money and engineering could do to the ocean. That’s the whole point of the object, and Cameron understood that the object was making a claim before he put anybody on it.
So the sinking isn’t a set piece dropped into a romance. It’s the counter-argument. The film spends its first hour establishing a floating demonstration of hierarchy, with the deck you’re allowed on determined by what you paid, and then it tips the demonstration into the Atlantic and asks who gets a lifeboat.
That’s why the class material survives being broad. It’s broad, and it’s also load-bearing. Rose’s engagement is a transaction and the film says so plainly. Jack has nothing except the ability to see her as a person, which the film treats as the only meaningful currency aboard.
The setting carries an argument or it’s scenery. Cameron chose a location that was already making a statement about wealth and nature before a single character walked onto it, and then let the plot demolish that statement. When you place a story, ask what the place asserts. If the answer is nothing, the place is doing no work and you could set it anywhere.
The practical application is to list out. A reader cannot total up abstractions, so broad sacrifice produces nothing at the payoff. Cameron gives you this trip, that college place, this exact sum on this exact day, and the audience keeps a ledger without being asked to.
Go through your manuscript and find every place a character has given something up. If the description is categorical, a dream, an ambition, a chance, replace it with the specific object, the date and the amount. The final release is worth exactly the sum of what the reader has been able to count.
Making a Known Ending Land
Every person watching knows the ship sinks, knows it happens on the maiden voyage, knows roughly fifteen hundred people die. There’s no suspense available about the outcome.
Cameron replaces suspense with specificity. The audience cannot be surprised by the sinking so he makes them watch it happen to particular people in a particular order. The band. The old couple holding each other in the flooding cabin. The designer setting the clock. The mother telling her children a story while the water rises.
None of those vignettes affects the plot. All of them are the point. The disaster isn’t one event, it’s fifteen hundred individual ones, and the film shows you enough of them that the number stops being a number.
The framing device does the same job from the other direction. Old Rose is alive, so we know she survives, which removes the last available uncertainty. What it adds is that we’re watching a memory, and memory is where the film ends up living.
When your reader knows the outcome, and in memoir and historical work they usually do, suspense isn’t available and chasing it wastes the book. Specificity is available. Show the thing happening to somebody in particular, at a particular hour, doing a particular small task. A statistic is a fact. A woman finishing a story she knows her children won’t survive is an experience.
The technique is to give both readings the same weight in every scene. Anything staged to be misread once will read as a cheat on the second pass, and the second pass is where an ambiguous structure either pays off or exposes itself.
The check is to list your scenes and describe each one twice, once under each interpretation. Any scene that only makes sense under one is load-bearing for a trick instead of a structure, and it’ll have to be rewritten or cut.
The Door
There was room on the door. This has been argued, tested, filmed by MythBusters, and eventually conceded by Cameron himself, who has said the script required Jack to die.
That answer is honest and it’s also the problem. A film isn’t permitted to require a death that the audience can see is avoidable while the camera holds on a piece of debris large enough for two. The audience is doing arithmetic during the most emotional scene in the picture, and arithmetic is fatal to feeling.
The fix was available and cheap. Make the door visibly too small, or damage it, or have Jack try and fail, or put him in the water before the debris appears. Any of those preserves the death and removes the objection. Cameron did none of them, and thirty years of argument is the result.
It’s the most visible flaw in a film with several, and it’s the only one that costs the film its ending.
If your plot requires something, the reader must not be able to see the alternative. A death, a betrayal, a decision that drives the third act, all of them work only if the character had no other move available. Close the exits before you need them shut. An audience that spots the unlocked door stops grieving and starts arguing, and you’ll never get them back.
The rule is simple and constantly broken: if the reader can see the alternative, the necessity is gone. Close the exits before you need them shut, and close them physically instead of emotionally, because a reader will accept that a door is locked and won’t accept that a character forgot to try it.
Audit every plot-critical death, betrayal and separation in your draft by asking what a competent, motivated person would have done instead. If you can answer, the reader can too, and they’ll be answering it during the scene you most need them to feel.
What the Money Bought
Two hundred million dollars, which at the time was the largest budget ever spent, and the striking thing is where it went. Not on spectacle for its own sake but on being able to shoot the ship as a real place.
Cameron built it at ninety percent scale and flooded it for real. That means actors are reacting to actual cold and actual water at an actual angle, and the difference is legible on screen in a way that survives every advance in effects since.
The dives to the wreck matter for the same reason. The film opens with real footage of the actual ship on the sea floor, which quietly establishes that this happened before it asks you to care about people who didn’t.
The research that shows is rarely the research that matters. What the budget bought here wasn’t visible accuracy but the ability to stage scenes in conditions close to the real thing, which produced performances that couldn’t have been faked. When you research, look for the detail that changes how a scene has to be written, not the detail you can display.
The general principle is that research should change what happens, not decorate it. Cameron’s money bought conditions that produced performances, which is research expressed as staging instead of as detail.
So when you research, sort your findings into two piles: things you now know, and things that alter a scene. The first pile is a temptation and mostly belongs in a notebook. The second is the only part that reaches the page usefully.
What Is Dated
The dialogue creaks in places and the villain is a villain in the mustache-twirling sense. Billy Zane’s Cal has no interior life and the film doesn’t want him to have one.
The class critique is a century of history compressed into a shorthand that wouldn’t survive a paragraph of scrutiny. Steerage is warm and authentic, first class is cold and cruel, and the actual Edwardian arrangement was considerably more complicated than that.
And the romance is between two types more than two people. Jack is decency without circumstances. Rose is a woman discovering she’s a self. Neither is complicated and the film moves too fast for them to become so.
That the picture works despite all of this is the interesting fact about it. It works because the ship is doing the thematic labor while the leads carry the feeling, and the division holds.
What the Framing Device Does
The film is told by a hundred-year-old woman to a salvage crew looking for a diamond, and the device is doing more than people credit.
It removes the last available suspense. We know Rose survives, so the two hours aren’t about whether she lives, which frees the film to be about what the experience cost and what she did with it afterward.
It also makes the film a memory, and memories are permitted to be heightened. The romance is impossibly compressed, the villain is a cartoon, the class material is broad, and all of it is defensible as the shape an old woman has given her own past over eighty-four years.
Gloria Stuart was eighty-six when she played the older Rose and was nominated for it, and she’s the reason the frame works. She plays a woman who has clearly told this story before and is choosing what to include.
The salvage crew are the other half of the device. Bill Paxton’s Lovett has spent years looking for a diamond and is confronted with the fact that the wreck is a grave, and his change of position is the film telling the audience how to feel about spectacle.
The last scene, where she drops the diamond into the ocean without telling anybody, is her keeping possession of the story. She’s given them the account and none of the evidence.
A frame narrator who’s remembering instead of reporting buys you permission to heighten, and it’s the cheapest available defense against the accusation that your material is too neat. Titanic’s romance is impossibly compressed and its villain is a cartoon, and both are consistent with an eighty-four-year-old account.
The requirement is that the narrator must be visibly shaping. Old Rose pauses, declines to answer a question, and chooses what to say next, and those moments tell the reader that what follows is a version. Without them the frame is just an introduction and the heightening reads as authorial clumsiness.
The test is to ask whether your framing narrator ever withholds anything. If they simply relay events in order with full access, you’ve gained nothing from the frame and paid for it in distance. Give them one thing they won’t say, and the entire account becomes something the reader has to weigh.
Should you watch Titanic?
Yes, and it holds up considerably better than its 1997 backlash suggested. Expect three hours, a cartoon villain and a door you’ll argue about. What you get is the hardest thing in narrative done properly: an ending everybody knows, landing anyway.
Frequently Asked Questions
This film appears on 7 of my film lists