Judy (2019) Review
5/10
Official trailer
A woman and two children are turned away from a hotel in Los Angeles for an unpaid bill. She has no money, no home and a custody hearing coming. She takes a five-week engagement in London because it’s the only work anybody will insure her for.
The 5 is for a competent picture built entirely around a performance, with a structure that’s one good idea and doesn’t develop it. Everything outside the Talk of the Town sequences is functional and everything inside them is the reason the film exists.
Judy leaves me in two minds. I admire what Zellweger does in almost every frame, and I keep wishing the film around her had half her nerve.
The argument on this page is about the award instead of the work. Zellweger’s Garland is a real achievement of technique, and the two performances it beat were doing harder things with less to hold onto.
People usually complain that the film is sentimental. The sentimentality is the least of its problems. The failure is that the picture never takes a position on its own subject, and a film about what a studio did to a teenager needs one.
The Transformation and Its Ceiling
Zellweger studied the footage, built the posture from it, and reproduced a physical vocabulary that’s on film and available to check against. The physical work is precise. A slight forward carriage of the head, hands that never settle, a habit of touching her own upper arm, and a way of arriving at a microphone that’s half performance and half support.
What she gets right beyond the mimicry is the exhaustion underneath the projection. Garland at forty-six could still produce the voice and the timing on demand, and the performance shows what producing them on demand was costing by the end of each number. The limit is that it remains an impersonation of somebody documented. Every choice can be measured against a recording. That’s what makes it legible to an audience and a voting body, and also what caps what it can discover.
There’s one sequence where the film gets past that. She’s alone in a hotel suite, isn’t performing for anybody, can’t sleep, and the scene has no reference footage behind it because nobody filmed Judy Garland alone in a room. That scene is what the whole film should have been made of. It’s four minutes long, arrives about seventy minutes in, and the picture returns immediately to material that’s a documented original to lean on.
The prosthetic work is also considerable and largely invisible. That’s worth crediting. Dental appliances and nose work instead of a mask, and the face stays mobile enough to act with. That’s the hardest part of that kind of build.
An impersonation of somebody documented is assessable and a portrait of somebody unwitnessed isn’t. Readers and juries reward the first because they can check it. The trap is that checkability caps discovery. If every choice can be measured against a source, nothing in the performance can surprise anybody who knows the source.
Find the scenes nobody recorded. A real person alone in a room is where a portrait stops being a reproduction, and this film locates exactly one of them. If you write a biographical subject, list what was never observed. That list is your material and everything else is transcription.
Five Weeks at the Talk of the Town
The engagement is the film’s frame and it’s a good one. A fixed run, a contract she can’t break, a nightly obligation, and a body that’s not going to manage all of it.
The structure that follows is a countdown and the film uses it loosely. Nights go well, then badly, then not at all, and the audience is given a run of performances of declining reliability instead of a plot.
Jessie Buckley plays the assistant assigned to get her on stage and the part is the film’s most useful invention. Rosalyn has a job to do, isn’t a friend, becomes something close to one, and the relationship is entirely built out of logistics. The best material in that strand is practical. Getting her out of a suite, into a car, through a corridor, onto a mark, and the film treats each of those as a separate problem with its own solution on any given night.
Where the structure fails is that nothing accumulates. Each night resets, the same crisis recurs with slightly different details, and by the fourth the audience has learned the pattern and is waiting instead of watching.
The comparison is the other entry on this page that’s the same fault. Crash reverses a position on schedule and this film collapses a performance on schedule, and in both cases the repetition teaches the audience the shape early.
The audience in the room is also underused. A London supper club in 1968 watching a woman they came to see fail is an uncomfortable situation, and the film treats them as reaction shots. I’d have given that crowd a face or two, because their disappointment is half the horror of those nights.
The Flashbacks Are the Main Subject
About a fifth of the film is 1938 on the Wizard of Oz set, with a teenage Garland managed, medicated and instructed by the studio.
The material is the strongest idea in the picture. Louis B. Mayer explains what she’s worth and what she owes, a matron administers pills on a schedule, and the film establishes that the adult in the London hotel was manufactured.
Darci Shaw plays the young Garland and the sequences are shot flatly and coldly. That’s the correct instinct. The studio is a workplace with a production schedule and the film declines to make it gothic. The pill scene is the one that works. She’s handed something to sleep and something to wake, is told both are necessary, is sixteen, and nobody in the room is behaving unusually by the standards of the room. What the film won’t do is connect the two halves with any force. The flashbacks are placed as explanation, arrive when the present-tense material needs a reason, and the picture never asks what anybody owes her now.
A version of this film that stayed in 1938 and ended at the premiere would have been considerably better and considerably harder to finance. The material that explains the subject is more interesting than the subject’s decline. I’d watch that version tomorrow. The pill scene rattles me more than all the London collapses put together, and it’s the only stretch of the film where I feel angry on her behalf. The one place the two halves meet properly is a line about being asked whether she was happy. It arrives late, is delivered quietly, and is the only moment the film permits itself an argument about what was done to her.
When your flashback material is better than your present tense, that’s information about which story you chose. Explanation is usually more interesting than consequence and almost nobody structures it that way. The commercial reason is that decline has a star part in it and formation doesn’t.
Film the institution as a workplace. A studio with a production schedule and a matron with a clipboard is worse than any villain, because nobody in the room thinks they’re doing anything unusual.
Ask what your work is arguing about the harm it depicts. This film shows the manufacture and then declines to say anything about who was responsible. That’s the difference between a subject and a position.
The Singing
Zellweger sings live on camera throughout and the decision is the right one even though the results are uneven.
Its defense is that a dubbed vocal removes the thing the film is about. Garland’s instrument at that point was damaged, unreliable, and still capable, and a clean studio track would have been a lie about the last year of her working life.
What live recording delivers is the strain. Notes that don’t arrive, a breath taken in the wrong place, a phrase pushed through instead of sung, and none of that survives a dubbing session.
Zellweger isn’t a comparable singer and the film is honest about it by not comparing. There’s no archive footage of the real vocal anywhere in the picture. That’s a deliberate omission and the only way the approach could work.
The final number is the film’s most effective sequence and it works because the singing fails. She can’t finish it, the room takes it over, and the audience in the club completes the song while she stands there.
That’s a good ending and it’s also the film’s whole method in miniature. The picture is at its best when it stops trying to reproduce a performance and shows a room full of people who wanted one.
The song selection is also conservative and the film is better for it. Six or seven standards everybody knows, staged plainly, with no attempt to reframe any of them as commentary on the plot.
Johansson and Theron
Best Actress went to Zellweger. Scarlett Johansson was nominated for Marriage Story, Charlize Theron for Bombshell, Saoirse Ronan for Little Women and Cynthia Erivo for Harriet. The case for Johansson is on this site. Nicole is a woman inside a legal process, has no transformation to perform, and every scene is ordinary behavior under pressure. That’s the harder thing to make visible. Theron’s case is the one people make less and it’s strong. Megyn Kelly required a physical build, a particular vocal placement, and a character the audience isn’t invited to like, and the performance holds a position the film keeps testing. The pattern is the one this page documents in every acting category. A documented original produces an assessable performance, ordinary behavior doesn’t, and the assessable version wins.
The career narrative also mattered and it’s worth naming. Zellweger had been largely absent for six years, this was a return, and a voting body responds to a comeback in a way that’s nothing to do with any individual scene.
The win is a real achievement of craft. My objection is that two performances that year were doing more with less scaffolding, and if the vote were mine the statue would have gone to Johansson. The film’s own awards profile supports the reading. Two nominations, one win, nothing for direction or screenplay or anything else. That’s the signature of a picture the membership regarded as a container for a performance.
A comeback is a narrative a voting body responds to and it works independently of the work. Nothing about that is in a performer’s control and it belongs in any honest account of a result. Two nominations and one acting win is the clearest institutional statement available. The membership assessed the performance and had no view about the film.
A performance with no reference material is competing at a structural disadvantage, permanently, in every award in every field. Write the version with less scaffolding anyway. It’s the harder work and it’s the one that’s still discussed when the transformation has stopped being notable.
Should you watch Judy?
Watch it for the last ten minutes and the 1938 material. I give it a 5 because Zellweger’s technique and that final song earn real respect, and the rest of the picture keeps repeating the same bad night without ever deciding who did this to her. The ending, with the room finishing the song while she stands there, is the part that nearly talked me into a higher number.
