Touch of Evil (1958) ***

Re-watching this film at the cinema last night, it occurred to me that there are directors like George Cukor, for whom technique is the art of the unobtrusive, defined purely as the most effective means by which the story can be told. Then there are the directors like Hitchcock for whom technique is the point, and storytelling is the art of finding excuses to display it. But few directors enjoyed Hitchcock's luxury of being able to select and tailor material specifically for the technical opportunities it would allow him. Most took what they could get and did what they could with it.
Orson Welles was an extreme example of the latter, a man who - despite his protestations to the contrary - loved dazzling the audience with audacious and innovative technique, but rarely got the opportunity to do so organically after Citizen Kane.
The inevitable result is a film like Touch of Evil, a cop thriller for which Welles was hired as an actor but upgraded to (re)writer/director at the instigation of Charlton Heston, where we see two enterprises running simultaneously - a narrative, about a Mexican detective investigating a crime in a border town who discovers that a respected American police captain is planting evidence to frame suspects - and a presentation, in which Welles crams in as much style, showmanship and experimentation as the routine structure will hold.
It's not seamless, the way that technique and content are seamless in the best Hitchcock or Cukor movies. Parts of it are messy, and the match is imperfect. The half-dozen or so most celebrated displays of directorial bravura sit adjacent to the film itself, rather than rise with it in the same dish.

Add to this the customary myth-making: Welles went off chasing his next project after assembling a rough cut, then was horrified to discover that in his absence Universal had re-edited the footage into a vague approximation of the film they hired him to make in the first place. But there was no conspiracy against Welles at the studio, as legend would have us believe. The only reason why he wasn't closely consulted in the post-production process is because nobody knew where to find him.
As a result, there are now three versions available: Universal's release cut, a longer one incorporating cut material but also retaining the extra scenes re-shot without Welles's involvement, and a restoration made in accordance with the changes requested by Welles in a letter he wrote to Universal after seeing their version, which is as close as possible to being a director's cut without the director actually being involved.
The latter is the most commonly available today, and while it is a better film than the Universal cut in a number of subtle ways, there's really not all that much in it. All available versions are good, share the same faults and - crucially - include all the showstopping Welles moments that have guaranteed the film's critical longevity.
The only really distinguishing feature of the Universal cut, apart from a very few unnecessary extra scenes designed to smooth out the plot, is their decision to run titles and theme music over the opening continuous take tracking shot. (Though even here, there might be a case for arguing that this helps integrate this technical display a little more organically into the subject. And I personally find the new edit - which begins without any kind of opening title at all - distractingly incomplete.)
So there should really be two reviews here, for two films.
As a chance for Welles, ten years after Lady From Shanghai, to show that he still has what it takes to turn a basic Hollywood thriller into a dazzlingly personal display of artful composition and presentational innovation: a certain masterpiece, fascinating from start to finish.
Four stars.
As a Hollywood thriller: a good, seedy, downbeat tale of corruption, with unusual and impressive location photography, excellent acting cameos and mordant dialogue. (That Welles is a superb director of actors is often forgotten, but both Charlton Heston and Janet Leigh have rarely been as relaxed and confident, and it really does feel like they're married to each other.) But the film is too long, key narrative points are needlessly obscured, and - in all available versions - the scenes with Dennis Weaver as the eccentric motel night man go on far too long. (In particular, his last appearance, with Heston at the now abandoned motel, is magnificently shot and lit, but seriously disrupts the tension of the film's final act.) Plus the title doesn't mean anything.
Two stars.
How much does it matter that these two films never quite coalesce? Up to you.