I made “The Marriages of Laurel Dallas,” my comparative study (above) of the cinematic build up to two tear-jerking moments—one long enshrined, and much argued over, in maternal melodrama studies—shortly after I saw the silent version of Stella Dallas for the first time. I had always assumed that the two earliest film adaptations of Olive Higgins Prouty's 1922 novel, from which the sequences in my video are drawn, would be quite different from one another in their mise-en-scène; I knew, at least, that they were made on either side of the divide between silent and sound cinema. I was also familiar with the view of film historian Christian Viviani, expressed in a footnote to his chapter “Who is Without Sin? The Maternal Melodrama in American Film, 1930-39” (in Gledhill’s Home is Where the Heart Is): “King Vidor’s  film follows quite faithfully the plot of Henry King’s  version. Only their very different approaches help distinguish between the two films.”4
It’s hardly unexpected, of course, that the two plots should be similar. The films are relatively faithful adaptations of the same novel, not just key members of the same cinematic genre, the 1925 film, in Viviani’s assessment, “an initial milestone” that played a “precursor role” in the “maternal melo;”5 the heroine of the second film “an archetype”6 of maternal decisiveness and energy, as well as suffering, as played by Barbara Stanwyck. Like Viviani, in her rich and fascinating 2011 comparison of the novel with King and Vidor’s films, Diane Stevenson7 also particularly flags up the differences in approach between these three versions of the Stella Dallas story. So, when I finally saw the silent adaptation recently, I don’t think I was prepared for what I then experienced as uncanny similarities between the endings of the two films. It was this affective experience of the analogous narrative and aesthetic choices of the two endings that was especially compelling to me, hence my video about (at one and the same time) it and them.
It is the silent era adaptation that first departs from the source novel in its climactic organisation around Laurel Dallas’s wedding (the novel concludes, instead, with Laurel’s formal presentation to New York society, even though this is witnessed through a window by her mother standing outside on the street before she is moved on by a policeman). The final scenes of the 1937 film not only emulate many of the staging and framing choices of the earlier film; in a number of important details they potentially prompt our active recall of them, and in so doing (as the later film also does elsewhere in its diegesis) they bespeak the pre-existence of the 1925 film in eloquent and intriguing ways. Perhaps most fascinatingly the second adaptation swerves away from the novel and King’s film in its portrayal, in an early scene, of Stella at the cinema, entranced while watching a romantic silent film, which, one might fancy, could even be (although isn’t) a scene from the first film. In all these ways, the 1937 film may be considered to be more aware about its cinematic precursor than much discussion of Stella Dallas in Cinema Studies has been, for it is true that these films are not equally prominent in, or even equally remembered by our discipline. Vidor and Stanwyck’s Stella Dallas went on to become an almost transcendent exemplar in theories of melodramatic pathos, in large part, because of its final scenes, as in Thomas Elsaesser’s renowned study:
Pathos results from non-communication or, silence made eloquent—[…] a mother watching her daughter’s wedding from afar (Barbara Stanwyck in Stella Dallas) […]—where highly emotional situations are underplayed to present an ironic discontinuity of feeling or a qualitative difference in intensity, usually visualised in terms of spatial distance and separation.8
The 1937 adaptation and its ending were also taken up as a touchstone in feminist critical debates about the cinematic forms of maternal melodrama as well as about the complex identificatory possibilities of the female film spectator.9 And the dénouement of the later film further became one of the markers of meta-cinematic consciousness in reflections on 1930s Hollywood: for example, as Viviani notes, “The framing of [Vidor’s final] scene, its lighting, even the form of the bay window where the wedding ceremony is enacted, evoke in a troubling way, the cinematic spectacle.”10
My surprising (and very pleasurable) experience of watching the silent film version (it is very beautiful) also shook my own deep investment in the specialness of King Vidor’s film and its final sequences. The 1937 movie, which I had loved when I saw it on television a number of times as a teenager, was one of the first films I chose to teach in a mid 1990s Masters class on feminist film theory, very early on in my academic career. I had my own VHS copy of it, while the King film was not so easily available, if it all. I was also fortunate to have been born into the generation below that of the truly great and pioneering feminist film scholars who produced so much understanding about this movie, its ending and its centrality to the genres of the “Woman’s film” and the maternal melodrama. Their published theorising was, as a result, easily available to me. So the ending became one of the two main sequences I could focus on in seminars, relying on their work as generative set reading. In short, the silent film version of Stella Dallas just didn’t stand a chance, for me, in the attentional or canonical economies of Film Studies at that point, either in terms of its salience or its accessibility.
My “silent” video remix (above) came from the simple desire to experience the two film endings—a beloved one to be re-visited and a previously unknown one to be newly explored— together in a live synchronous comparison, now that both had become available to me in digitised versions. I had experimented with this kind of videographic comparison before, and indeed wrote about it in an earlier issue of Mediascape. Unlike those experiments, a simple split screen or sequential montage didn’t seem to work with this material: those techniques kept everything too separated, too academic, too inert. I needed to play in a more complex way with proximity and distance videographically, on this occasion. So the comparative framework that I found myself adopting was one of a synchronous, slightly over-lapping, side-by-side placement, with the silent film higher in the frame, and the sound film lower. The 1925 footage was made more translucent for its partial superimposition over the 1937 sequences. The later film was slightly slowed in motion and newly colorized to correspond with the presentation of the silent film.
As I wrote of the series of related experiments in my earlier Mediascape essay:
With their precise juxtapositions of film material, which unfold in […] space and time, video essays, or assemblages, like the one above, can introduce us to the “unconscious optics” of particular instances of intertextuality. […]
As well as an exposure to audiovisual argumentation (involving selection of evidence, montage and mise en scene, titling, sound editing and other creative effects), they offer an active viewing process, one of live co-research, or participant observation. Unlike written texts, they don't have to remove themselves from film-specific forms of meaning production to have their knowledge effects on us. And we can feel, as well as know about, the comparisons these videos enact.11
This kind of work is not only comparative, of course, but also combinative, and thus creative, especially in this case, given the added music (credited below) that doesn’t come from either version of Stella Dallas. Unlike a good deal of the existing critical work on the films, the video’s aims are not directly hermeneutic but meditative, exploratory, experiential. It plays with and expands upon the kinds of generic and intertextual repetitions and differences that the films themselves perform, between each other and the source novel. The resulting video also acts out and multiplies the spectatorial analogies of the final scenes of both films, with characters from each “watching” scenes in the other and at least appearing to seek an even more complex and trans-generational exchange of looks than in either film on its own. In a similar way, the video proliferates external spectatorial positions, or entry points, and perhaps further multiplies identificatory possibilities, too. The latter is just one of the most obvious points of contact that it has with the theorising of those feminists and other scholars who helped to inspire, as much as the original films, the associative and affective work of memory and (re-)creation it performs.
The video embedded above is a comparison of the final scenes of two cinematic ‘Maternal Melodrama’ adaptations of Olive Higgins Prouty's 1922 novel Stella Dallas: the 1925 version directed by Henry King, starring Belle Bennett as Stella Dallas, Lois Moran as Laurel Dallas, Alice Joyce as Helen Morrison and Ronald Colman as Stephen Dallas; and King Vidor's 1937 version starring Barbara Stanwyck as Stella (Martin) Dallas, John Boles as Stephen Dallas, Anne Shirley as Laurel Dallas and Barbara O'Neil as Helen Morrison.
The video was first screened at the Maternal Melodrama Symposium, University of Kent, UK, on June 3, 2014. Thanks to Pam Cook, Elizabeth Cowie, Tamar Jeffers McDonald and others in the audience for their valuable comments.
The music used in “The Marriages of Laurel Dallas” is adapted from the track 'Sweet Tender' by Jared C. Balogh, shared at the Free Music Archive under an Attribution-NonCommercial-ShareAlike 3.0 International License: http://freemusicarchive.org/music/Balogh/Lunar_Elegance/SWEET_TENDER