Fanfare as Fulcrum: A Pivotal Event in Max Steiner’s Theme for Warner Brothers

[0.1] Shortly after signing a long-term contract with Warner Brothers (WB) in 1937, pioneering film composer Max Steiner created a signature fanfare for the studio, using bold brass arpeggiations and whirling strings to introduce the newly-designed WB shield (see Example 1).⁽¹⁾ Steiner’s fanfare served as the introduction to most of the studio’s films for the next two decades, spanning much of the Classic Hollywood era. Many studios had such fanfares—another well-known example is Alfred Newman’s fanfare for 20th Century Fox (composed in 1933).⁽²⁾ But what makes Steiner’s fanfare distinctive and particularly interesting is that it was composed without a clearly defined ending: it continues directly into the opening title sequence without pause.⁽³⁾ And although it always begins in the same way, it frequently resolves in unexpected ways, rarely repeating resolution patterns from previous films.

[0.2] In examining Steiner’s opening title sequences as part of a larger corpus study project (Lyon and Yorgason 2020), we found that Steiner experimented with various harmonic and melodic possibilities in resolving this opening fanfare and that these diverse resolutions seemed to set the overall tone for the film.⁽⁴⁾ We also observed that Steiner frequently interspersed a short transition between this resolution and the arrival of the main title music (which was often in a different key, meter, or tempo). These initial observations led us to pursue a corpus study to determine the following: (1) How thoroughly diverse are the fanfare resolutions in Steiner’s opening titles? (2) How effectively do these resolutions convey the tone of the film? And (3) What role does the subsequent transition play in the opening title sequence?

[0.3] The results of this study are outlined below, with complete data being provided in three appendices. We begin with an overview of the history of the WB fanfare. Then, we describe the corpus study itself, outlining our research methods and detailing the findings for each of our three research questions: variety of resolutions, communication of tone, and the function of the transition. Following this is a discussion of some notable outliers in the study and a brief summary of our overall findings.

History of the WB Fanfare

[1.1] The first version of the Warner Bros. fanfare appeared in the film Tovarich, which was released on Christmas Day, 1937.⁽⁵⁾ Example 2a provides Steiner’s original sketch for the fanfare from Tovarich, and the accompanying video provides the music from the film.⁽⁶⁾ The sketch clearly illustrates that he associated this theme with the “W.B. Shield.”⁽⁷⁾ The transcription of Steiner’s sketch in Example 2b highlights the principal aspects of the fanfare: a quick upbeat gesture, triadic brass arpeggiations (in triplets), ascending scales and trills in the strings, and a chromatically-descending bass line.⁽⁸⁾

[1.2] Steiner likely created the fanfare for Tovarich by using bits and pieces from the opening themes of two other films: A Star is Born (see Example 3) and First Lady (see Example 4). Both of these films were released in mid-1937, shortly before he began working on Tovarich. In A Star is Born (composed for Selznick Pictures), we see evidence of the first half of the WB fanfare, with the melodic content of the first measure (bracketed in Example 3) mapping onto m. 1 of Tovarich. Quick upbeat gestures and triadic arpeggiation figures are also introduced here. The second half of the WB fanfare comes from First Lady, with the bracketed arpeggiations mapping onto m. 2 of Tovarich melodically. First Lady also adds the chromatically-descending bass line and ascending scalar string gestures.

[1.3] Unfortunately, we don’t know the impetus for the creation of the fanfare itself. John Morgan (personal communication, Sept. 26, 2018) reports that Steiner only thought of using it as the official Warner Bros. fanfare after having composed it for Tovarich, and he was encouraged by his long-time orchestrator, Hugo Friedhofer, to try using it again in his next film.⁽⁹⁾ But his colleagues quickly adopted it—long before Steiner himself would complete another film (see Example 5). Just two weeks after Tovarich was released, the fanfare was used by Adolph Deutsch in Swing Your Lady (released on January 8, 1938), and Ray Heindorf used it in Hollywood Hotel just one week after that (January 15). Heinz Roemheld’s first film of the year, Slight Case of Murder (March 5, 1938), also used the Steiner fanfare.⁽¹⁰⁾ Whether or not the fanfare was something the studio had asked Steiner to create, they clearly received it appreciatively, since for the next two decades the fanfare was used by Warner Bros. composers to open the majority of the films scored for the studio.

[1.4] The C major setting of the fanfare in Gold is Where You Find It (Example 6a), Steiner’s next film for Warner Bros. (released in early 1938), became the “canonical” version to which he most frequently referred.⁽¹¹⁾ Other than the transposition from B♭ to C major, the two versions (Tovarich and Gold) are essentially the same until the downbeat of m. 3, where they resolve quite differently (compare Example 6b with 2b above).⁽¹²⁾ In Steiner’s subsequent film sketches, the fanfare is generally not written out in detail. In most cases, he either writes out just the melodic line (as seen in Example 7, from the sketch for Angels with Dirty Faces) or simply provides two blank measures and includes a note like this: “from M. TITLE ‘Gold is where you find it’ Orig Key” (as seen in Example 8, from the sketch for The Adventures of Mark Twain).⁽¹³⁾ Sometimes a brief pickup gesture into the music that follows is included as well, which may or may not differ from the reference model.

[1.5] Example 9 provides an overview of Steiner’s use of the WB Fanfare, from Tovarich in 1937 to its final appearance in Battle Cry in 1955. Other than a few war-related films in the forties (which more typically began with patriotic music)⁽¹⁴⁾ and three other exceptions,⁽¹⁵⁾ Steiner used the fanfare exclusively until 1948, at which point he started to substitute other newly-composed fanfares and overtures, beginning with his main title music for Treasure of the Sierra Madre in 1948.⁽¹⁶⁾ By the mid-’50s, the Steiner fanfare had disappeared. Some of the possible reasons for its demise were (1) changes to the WB shield logo itself in 1948, perhaps spurring accompanying musical changes;⁽¹⁷⁾ (2) the emergence of a younger generation of Hollywood film composers with new approaches to scoring (such as Alex North’s jazz-influenced score for A Streetcar Named Desire in 1951); and (3) the growing trend of using popular songs to open a film (such as Dimitri Tiomkin’s “Do Not Forsake Me” for High Noon in 1952). Or, perhaps, (4) Steiner had simply run out of interesting and new ways of resolving the fanfare into the main title music, and so he decided to move on to something else.

[1.6] In 1985, the classic WB shield returned, and for a time, Steiner’s WB Fanfare returned along with it as the “Warner Home Video” theme (see Example 10). This new, highly-synthesized, somewhat cheesy version of the fanfare ended rather predictably on a half cadence without any overlapping connection to the main theme. Notably, this perhaps too-obvious resolution to a half cadence on G was a solution that Steiner never used.⁽¹⁸⁾

The Corpus Study

[2.1] Corpus studies have become increasingly prevalent in music-theoretical research during the past decade, with corpora ranging from Rolling Stone’s “500 Greatest Songs of All Time” (De Clerq and Temperley 2011) to large collections of songs by The Beatles, The Stones, Hendrix, and Led Zeppelin (Biamonte 2014) to the keyword content of music theory journals (Duinker and Léveillé 2017) to the entire MIDI file contents of classicalarchives.com (White and Quinn 2016). The corpus study approach is invaluable when looking for larger patterns and trends or when dealing with large amounts of data. It has only recently been applied to film studies, with the most notable instance being Mark Richards’s study of phrase structure in the main themes of 482 Oscar-nominated film scores (2016).⁽¹⁹⁾

[2.2] The creation of a Max Steiner corpus is an ongoing project at Brigham Young University, incorporating data from 192 original film score sketches with 31,373 pages of hand-written music. Steiner was not only prolific (with over 350 film scores created between 1929 and 1965) but also highly influential. Buhler and Neumeyer note that “there is no question about the influence of Steiner’s practices. . . . By [1932], all composers in the major studios had assimilated the core of Steiner’s methods into the basic technique of film scoring” (Buhler and Neumeyer 2014, 38). Thus, the Steiner corpus is an invaluable resource in understanding major trends in film scoring during the Golden Age of Hollywood.

[2.3] In our corpus study, we looked at all eighty-eight films scored by Steiner that begin with the WB fanfare. The full list of films in our study is provided in Appendix A.⁽²⁰⁾ We found that the fanfare itself is rarely varied in any significant way. Except for a handful of exceptions (some of which will be discussed later), most of these openings refer back to Gold is Where You Find It,⁽²¹⁾ begin in C major,⁽²²⁾ and are played at a moderate tempo (see Examples 11 and 12).⁽²³⁾ Yet Steiner was able to use the same opening fanfare as a launching pad to a wide variety of themes in different keys and tempi and with quite different characters.

[2.4] The results of the three principal questions explored in our corpus study are presented in the sections that follow. In the first section, we demonstrate that although Steiner did not use every possible harmonic resolution for the WB Fanfare, he did actively attempt to find new ways to resolve it, and did so with a remarkable degree of thoroughness. We use Neo-Riemannian theory as well as various measures proposed by Frank Lehman (2013), Scott Murphy (2014), and Joti Rockwell (2009) to determine how diverse Steiner’s fanfare resolutions are. In the second section, we develop a simple formula to predict how the resolution of the fanfare might communicate the overall tone of a film, and we correlate this information with the film’s genre. In the third section, we show that the principal function of the transition following the fanfare was to delay the arrival of the main theme until the film’s title appeared on the screen.

Variety of Resolutions

[3.1] In the eighty-eight fanfare examples used in our corpus, the point of arrival (the moment at which the fanfare ends and the transition or main title music begins) typically occurs on the downbeat of m. 3. This pivotal event acts as a fulcrum in the opening title sequence, from which point of balance the music leans in a different direction tonally, texturally, temporally, and melodically. Here we will focus primarily on the variety of harmonic resolutions at the point of arrival, but will consider the effect of different melodic resolutions as well.

[3.2] For the purpose of making direct comparisons between the films in this study, we will consider all of them as though they had been composed in C major (transposing as needed). Each fanfare progresses from the initial C major triad to a B major triad just before the moment of resolution. From B major, Steiner resolves to all of the major and minor sonorities shaded pink in Example 13.⁽²⁴⁾ He seems to prefer resolutions to major sonorities (the outer ring of the circle), avoiding only the A major triad.⁽²⁵⁾ Along the inner ring of the circle, he resolves to just six of the twelve possible minor sonorities. Still, the overall coverage is remarkably broad. All of the harmonic resolutions in the corpus can be seen in the reductions provided in Appendix B (see m. 3 of each example—films that share a resolution are listed together).

[3.3] This resolution information can also be represented on a Tonnetz. The numbers on the Tonnetz in Example 14a represent the number of transformations in the corpus from B major (the blue triangle in the center) to other major and minor triads.⁽²⁶⁾ This diagram illustrates that Steiner clearly preferred some resolutions, with 10 instances resolving to E♭ major, 7 to B♭ major, 6 to D major, 5 to F major, and 4 to E♭ minor. Again we see that he favors resolutions to major sonorities (the upside-down triangles): 39 resolutions (75%) were to major triads, as opposed to 13 resolutions (25%) to minor triads. The diagram also shows that Steiner avoided almost all triads that had a B in them.⁽²⁷⁾ Perhaps the avoidance of B is the result of a desire to continue the chromatic motion from C through B to B♭, as occurs in the bass line of the original Gold model (see Example 6b). Indeed, Example 14a confirms that a slight majority of the resolutions (26 out of 49, or 53%) have a B♭ in them (see the resolutions surrounding “A♯” in the diagram). No other pitch is as likely to occur in a resolution.

[3.4] Since Example 14a includes just triadic resolutions, only 49 of the 88 resolutions in the corpus are represented. Example 14b is an attempt to remedy this by including resolutions to other chords as well.⁽²⁸⁾ Most noticeably different in Example 14b is the increased number of resolutions to C major (from 3 to 16), A minor (from 2 to 9), and G minor (from 2 to 8).⁽²⁹⁾ These now can be added to the list of Steiner’s preferred resolutions (along with E♭, B♭, D, and F major). Resolutions with B in them are still the least common (5%), and resolutions with B♭ in them are still the most common (39%), particularly if the 11 instances of C major with a B♭ seventh are included (51%).⁽³⁰⁾ Overall, there are fewer areas of the Tonnetz that are left untouched.

[3.5] Also included in Appendix B is information about the Neo-Riemannian transformations involved in each of these resolutions (shown above the triadic reduction following each progression).⁽³¹⁾ For example, the progression from B major to D♯ major in Tovarich (#1 in the appendix) is listed as LP (a leading-tone exchange followed by a parallel transformation) and the progression from B to D major in Dawn Patrol (#10 in the appendix) is listed as PR (a parallel followed by a relative transformation). In cases involving inter-cardinality (such as a triad resolving into a seventh), Jay Hook’s (2007) cross-type transformation (L’) is used.⁽³²⁾ For example, the progression from B major to C dominant in Gold Is Where You Find It (#2 in the appendix) is listed as NL’ (a near fifth transformation followed by a cross-type transformation that maps the triad onto a dominant seventh).

[3.6] Whenever the cardinality does not change (i.e., B major resolves into another triad), an additional label identifying the tonal-triadic progression class (TTPC) is included.⁽³³⁾ For example, the M4m label for The Oklahoma Kid (#11 in the appendix) indicates that a major triad resolves to a minor triad that is four semitones higher (B major moving to D♯ minor).

[3.7] The TTPC label is frequently paired with a chromatically modulating cadential resolution (CMCR) type. In Appendix B, the CMCR type is given in brackets following the TTPC. Lehman (2013) defines the CMCR (using music from Steiner’s Key Largo) as “a category of tonal motion that straddles the line between cadence and modulation” involving a dominant in one key moving to a tonic in a chromatically unrelated key.⁽³⁴⁾ This is a type of “sudden change, cleverly handled” ([5.3]). The fanfare resolutions in this corpus cannot always be understood as cadences, often landing on unstable harmonies that initiate transitional passages. But when they do resolve in a cadential manner, these cadences can almost always be described as CMCRs.⁽³⁵⁾

[3.8] Below each of the triadic reductions is a parsimonious voice-leading matrix (or P-matrix), which identifies all of the half- and whole-step motions required to move from B major to the sonority that follows.⁽³⁶⁾ And following each P-matrix is the total voice leading measure for the progression, or the total number of semitonal shifts required to get from the first sonority to the second. For example, in Crime School (#4 in Appendix B), 1 ascending semitone and 3 descending semitones give the total voice leading measure of “4”.⁽³⁷⁾ All of this information—whether it be the number of moves on a Tonnetz, the intervallic distance between roots, or the number of semitonal shifts—gives us some measure of the “distance” between the two sonorities involved.

[3.9] Example 15 provides a summary of the resolution types found within the corpus, ordered by resolution sonorities. Adjacent rows that are shaded in the same color bear certain similarities that might allow them to be grouped together. For example, the two rows grouped as #11 are enharmonically the same. Resolutions might differ only by added melodic dissonance (as with #5), or chordal inversions (see #19). In some cases, the only difference between two resolutions might be how the melodic line resolves. For example, one of the #9 resolutions features a LEAP UP in the melody while the other features a CT (common tone).⁽³⁸⁾ One might also choose to group together resolutions that differ only by the addition of a dissonant pedal tone (see the last two rows in #32, for instance). Lastly, there are four special cases to be considered (shaded yellow in Example 15). Each of these fanfares is harmonically extended such that its resolution at the point of arrival is no longer a B major sonority (see #54–57 in Appendix B). These can clearly be included as unique fanfare resolutions. However, if one’s focus is on the variety of ways Steiner resolves the fanfare from B major, then these would need to be excluded.

[3.10] Example 16 examines the percentage of unique resolutions based on these varying criteria. If we were to consider all of the rows in Example 15 as unique, there would be 56 different resolutions—or 64%, which is nearly two-thirds of the 88 films in the corpus.⁽³⁹⁾ As we eliminate resolutions that share similarities (such as resolutions that are enharmonically related, that differ only by melodic motion, etc.), the percentage correspondingly decreases. At the bottom end of the scale, when considering only harmonic resolutions that cannot be grouped through the elimination of dissonance, inversions, or dissonant pedals, 41% of the resolutions are still harmonically unique.⁽⁴⁰⁾

[3.11] Although enharmonic resolutions can clearly be ignored, we would argue that the other criteria listed in Example 16 have a more significant effect on the perception of a fanfare’s resolution and should not be eliminated. This leads us to choose the higher percentage of 61% uniqueness, which is remarkably high. The diversity of resolutions is particularly noticeable in the first decade of the fanfare. Considering only the fanfares composed between 1937 and 1947, 48 of these 62 resolved differently (77%, or over three-fourths). After 1948, Steiner tended to fall back on previous resolutions more often, even reusing thematic material from previous films on occasion. For example, the westerns South of St. Louis (1949) and Springfield Rifle (1952) both reuse the transitional theme from Virginia City (1940), which results in them all using fanfare resolution #2.⁽⁴¹⁾

[3.12] Returning to Example 15, there are other aspects that reveal Steiner’s penchant for variety. Of the 47 resolutions that can be described in terms of Neo-Riemannian transformations (“NRT” in column 3), 29 of them are unique (62%), and of the 31 resolutions that can be labeled with a tonal-triadic pitch class (“TTPC” in column 4), 16 are unique (52%). In fact, Steiner uses two-thirds of the 24 possible TTPCs. Notably missing is anything that resolves up a perfect fourth (there are no M5M or M5ms). Since this is the root motion in an authentic cadence, Steiner may have avoided it because it would have sounded too final for the opening title sequence.

[3.13] Column 5 in Example 15 provides the CMCR type for each resolution. Lehman (2013) identifies seven types of CMCRs (see his Example 25) but notes that they “come in diverse guises” ([5.5]), suggesting that there may be others. Accordingly, we have added two additional types from the Steiner corpus that meet the CMCR criteria (types “VII” and “VIII” in Example 17). These types resolve from a dominant to tonics in ♭VII and II♯. For example, Dawn Patrol (#10 in the appendix) resolves from B to D (V to ♭VII in E minor, or “Type VII”) and Santa Fe Trail (#22 in the appendix) resolves from B to F♯ (V to II♯ in E minor, or “Type VIII”). We have also included some CMCRs that resolve to minor tonics, by appending an “m” to the type label.⁽⁴²⁾ For example, Watch on the Rhine (#30 in the appendix) resolves from B to G minor (V to ♭iii in E minor, or “Type IIm”). Overall, Steiner uses all but two of Lehman’s CMCR types (representing 59% of the cadences), and creates six additional types of CMCRs (35% of the cadences, with two resolving to major tonics and four to minor tonics).⁽⁴³⁾

[3.14] Examining the number of semitonal shifts (column 6), we see that there is variety here as well, although Steiner’s resolutions predominantly involve three or four semitonal shifts.⁽⁴⁴⁾ In Rockwell’s (2009) study of parsimonious voice leading, he includes nine “types” (see “PVL Type” in column 7). Of these, the only one Steiner did not use was Type III.⁽⁴⁵⁾ Column 8 provides data regarding Steiner’s preferred inversions. Although he resolves to all inversions of triads and sevenths, he strongly prefers resolutions to the most unstable inversions at the point of arrival, which initiates the transitional section. Notably, 68% of all triadic resolutions are to a second inversion triad (36 out of 53) and 40% of all seventh resolutions are to a third inversion sonority (11 out of 27). The harmonic instability of the transition will be further discussed below.

[3.15] Moving beyond the fanfare, Steiner’s quest for variety continues. We discovered that from the initial key of C major, Steiner transitioned into main themes written in nearly every possible major and minor key (see Example 18). The yellow keys are those that appear literally, and the lighter yellow keys are additional keys that emerge when you transpose the fanfare to C major.⁽⁴⁶⁾ Perhaps nothing sums up Steiner’s desire for tonal and harmonic variety (and suggests that it was indeed a conscious one) than this example.

Communicating the Tone of the Film

[4.1] As we examined how various resolutions of the fanfare might help to communicate the overall tone of a film, we developed a theory. First, we hypothesized that the tone would be established as soon as possible after the opening fanfare. Certainly, Steiner would have been aware of the need to establish the tone of the film from the outset, and he would want to do so very soon after the studio fanfare. We then theorized that we might be able to predict the overall tone of the film by using just two pieces of information taken from the point of arrival of the fanfare: (1) the sonority type of the chord that the fanfare resolves to, and (2) the way that the melody resolves.

[4.2] Example 19 provides the values that we used to calculate the tone of the film based on these two pieces of information. These values were determined a priori and were chosen through a combination of intuition and experience (rather than referring to any pre-existing models).⁽⁴⁷⁾

[4.3] Melodically, the fanfare might resolve in a variety of ways:⁽⁴⁸⁾ it could end with a heroic leap upwards, a hopeful appoggiatura, an optimistic step up, a positive leading tone resolution, or a yearning retardation. All of these were assigned positive values ranging from +1 to +0.25. In other cases, the melody might resolve with an unexpected common tone resolution, a negative deflection downwards, or a rather tragic fall downwards. All of these were assigned negative values ranging from -0.25 to -1. The melodic resolutions for all of the fanfares in the corpus are provided at the end of the fanfare reductions in Appendix B.⁽⁴⁹⁾

[4.4] The sonority types were also given values ranging from the brightest and least dissonant sounds (major triads, dominant sevenths, and major sevenths) to the darkest and most dissonant sounds (minor sevenths, minor triads, and half-diminished sevenths).⁽⁵⁰⁾ As Example 19 illustrates, these values gravitate toward the extremes of +1 and -1, with no real “neutral” values given. To this measure, additional negative valuations were given for added dissonance (such as appoggiaturas, ninths, and elevenths) and dissonant pedal tones. The sonority types for each of the fanfare resolutions in the corpus are provided in Appendix B (listed just before the melodic resolution).⁽⁵¹⁾

[4.5] These two values (for the melodic resolution and the sonority type) were then combined, averaged, and multiplied by 100 to create a final range of values from 100 (representing the brightest overall tone) to -100 (the darkest overall tone).

[4.6] The results can be seen in Example 20, which provides a list of all of the films included in the study and their calculated tone at the point of arrival.⁽⁵²⁾ At the top of this list are the brightest films (those with positive values) and at the bottom are the darkest ones (those with negative values). Although there are a few interesting outliers in this list (which would likely only be evident to those familiar with the films in question),⁽⁵³⁾ the calculated tone at the point of arrival matches the overall tone of each film quite well.

[4.7] For each of the films in the study, we also collected the genre designations listed on IMDb.⁽⁵⁴⁾ We then correlated those genres with the calculated tone at the point of arrival for each film (again with a range from 100 to -100), as illustrated in Example 21. The chart shows remarkably good correlation with genre associations. Musicals, family films, biographies, sports films, and comedies were the brightest. Mysteries, horror films, thrillers, film noir, and crime films were deemed the darkest. We do not mean to suggest that the resolution of the fanfare can communicate the genre of the film by itself (providing as it does just a split-second of musical information)—only that there is good correlation between the tone established by the fanfare’s resolution and the general character of each film genre.

[4.8] Example 22 plots this same information onto a bubble chart using melodic resolution and sonority type as the two axes. The size of each bubble represents the number of instances in the study. Drama and romance are clearly the most common genres.⁽⁵⁵⁾ Note particularly how musicals (in the top right corner) and mysteries (in the bottom left corner) represent the polar opposites of bright and dark, whereas more neutral genres like “drama” and “action” gravitate towards the center. “Comedy” (in the bottom right corner) features bright sonority types, but unexpected resolutions (as one would expect from a comedy). Its polar opposite on the chart is “horror” (in the top left corner). “Sport” films (in the top center) were the most aspirational, featuring the most optimistic melodic resolutions of any genre.

[4.9] Example 23 provides an aural overview of these results with twelve selected fanfares from the corpus, ranging from one of the brightest—Daughters Courageous (1939), featuring a leap up to a major sonority with a calculated tone of 100—to one of the darkest—Key Largo (1948), featuring a common-tone resolution to a half-diminished seventh sonority over a dissonant pedal with a calculated tone of -100. As you listen to the fanfares in Example 23, note the gradual progression from light to dark as the melodic resolutions become less optimistic and the sonority types grow darker, and note the corresponding move from comedies and biographies at the top of the list to film noir and thrillers at the bottom.⁽⁵⁶⁾

[4.10] Our study also revealed some interesting details pertaining to specific genres. For example, in the romance genre, fanfares were more likely to resolve with a leap down, a deflection, or an appoggiatura than with a standard leading tone resolution.⁽⁵⁷⁾ Perhaps this represents the unpredictability or volatility of romance. Fanfares in crime films were significantly more likely to end with an unexpected common-tone resolution and were unlikely to feature a heroic upward leap. Perhaps this represents the desire to not heroicize criminals. Westerns typically featured heroic or positive rather than tragic or negative resolutions.⁽⁵⁸⁾ Films noir were significantly more likely to resolve with an unexpected common tone or a leap down than with a heroic leap up or step up. This reflects the dark, generally more pessimistic tone of film noir. War films, on the other hand, seemed to favor an optimistic tone, with upward leaps happening 45% of the time.⁽⁵⁹⁾

Function of the Transition

[5.1] In studying Steiner’s main title sequences, we found that the fanfare is often followed by a transitional passage that leads into the main theme. Our initial hypothesis was that the purpose of this transition would be to enable a convincing musical shift from the fanfare to the main title. But overall, we found that these transitions did not demonstrate clear tonal, temporal, or metrical directionality.⁽⁶⁰⁾ Steiner does not necessarily attempt to smooth out juxtapositions between keys, tempi, and meters in his opening title sequences—rather than using a transition for this purpose, he might just as likely jump directly to the main theme in its foreign key.⁽⁶¹⁾

[5.2] There is, however, a more consistent function for the transition: one related to visual cues rather than musical motivations. We found that whether or not a transition was used, the appearance of the film’s title on-screen coincided with the beginning of the main theme 80% of the time (in 70 of the 87 cases for which we had film evidence).⁽⁶²⁾ A typical opening title sequence might go like this:

1. The opening fanfare is played as the WB shield appears on the screen.
2. The names of the starring actors are shown on-screen during the transition (if any).
3. The film’s title appears on-screen, coordinated with the beginning of the main theme.

Example 24 illustrates this pattern with the opening title sequence from Virginia City (1940), outlining the visual elements, principal functions, and musical characteristics of each section.⁽⁶³⁾

[5.3] It is clear that Steiner wanted to coordinate the appearance of the film’s title with the first occurrence of the main theme. The principal purpose of the transition thus seems to be to delay the entrance of the main theme until the title of the film appears on the screen.⁽⁶⁴⁾ Steiner needed to pace the transition (both in his scoring and in his conducting) to match the visual structure of the opening titles given to him by the studio. This might explain why the lengths of the transition sections are so diverse, ranging from 3 beats to 40 beats in length.⁽⁶⁵⁾ In the most extreme of these cases—the 40-beat transition in Lightning Strikes Twice (1951)—the main title is delayed until 0:26 seconds into the film. Here, Steiner continues to build tension by using an ascending sequence that highlights the appearance of each the names of the four starring actors (followed by the punctuation of two lightning strikes!) before finally bringing in the main theme (see Example 25).

[5.4] Evidence from the corpus further corroborates the theory of visual coordination as a principal factor in composing transitional sections. 82% of the transitions (55 of 67) are coordinated with the appearance of the headlining stars’ names on-screen. On the flip-side, two-thirds of the films without transitions (14 of 21) lack any starring names before the main title appears (presumably because the lead actors were not A-list stars).⁽⁶⁶⁾ Finally, consider the opening sequences for Dodge City (1939) and San Antonio (1945). Both films use the same fanfare and main theme (in two different keys), but their transitions differ quite a bit, allowing the main theme in both cases to appear precisely as the title of the film appears on screen (compare these transitions in Example 26).

Notable Outliers

[6.1] Although in most cases the WB fanfare itself is not altered, Steiner did alter the fanfare through phrase extension and harmonic deviation in a few cases. Some of the fanfare resolutions also feature unexpected sound effects or involve dramatic transformations. Here we describe some of these outliers.

[6.2] First, we will examine a few instances in which Steiner extended the fanfare beyond its typical eight-beat length. The fanfare for The Amazing Dr. Clitterhouse (1938) is extended one measure by repeating motives from the fanfare, resolving only when the main title appears on the screen (see Example 27). The obsessive repetition here might reflect Dr. Clitterhouse’s unhealthy preoccupation with the study of crime.⁽⁶⁷⁾ In One Foot in Heaven (1941), the extension of the fanfare allows the music to expand harmonically as well, with the outer voices continuing in contrary motion from B major through E♭ minor and A♭ minor to a transcendent arrival on F major in the fourth measure (see transcription #55 in Appendix B and Example 28).⁽⁶⁸⁾

[6.3] Another means of subverting the resolution of the fanfare is to have it overlap with an unexpected sound effect. For example, the fanfare in The Bride Came C.O.D. (1941) resolves to what sound like wedding bells at the point of arrival. Steiner voices the instruments in clashing open fifths and adds ringing chimes to create this effect (see Example 29). In White Heat (1949), the downbeat gesture at the point of arrival mimics the sound of a train whistle. Steiner creates this effect with sliding parallel polychords: C and B major triads slide up a half step to D♭ and C. At the same time, we see a train approaching on-screen and hear it moving slowly along the track (see Example 30).⁽⁶⁹⁾

[6.4] One of the most remarkable transformations of Steiner’s fanfare occurs in the film Jim Thorpe - All-American (1951). Instead of referring back to Gold or Tovarich, Steiner writes out the entire fanfare here, recomposing it with energetic internal repetitions that propel the fanfare forward to the shouted cry of a college fight song at its climax. The fanfare as a whole is extended from two measures to seven-and-a-half measures and is played at the unusually quick tempo of 132 bpm (the outlier in Example 12), reflecting the speed and excitement of the racing athlete (see Example 31).

[6.5] In our final example, from The Great Lie (1941), Steiner had an opportunity to parody himself as a film composer. Although the movie does not begin with the WB Fanfare, later on (at 1:22:18) there is a “film within the film”—a home movie made by a father for his infant son. The music for the home movie is laden with clichés, quoting “Flying Trapeze” for scenes in an airplane, silent-era “villain music” for the meddling maid, “Sailor’s Hornpipe” for a toy boat in the tub, “How Dry I Am” for the drinking (milk) scene, “Turkey in the Straw” for a scene at the farm, and “Rockabye Baby” for the bedtime conclusion. At the beginning of the home movie, Steiner’s fanfare is played on what sounds like toy instruments, buzzy and shrill in tone (see Example 32).⁽⁷⁰⁾ It must have been fun for Steiner to be able to go over the top with this caricatured film score and to poke a bit of fun at the WB Fanfare as well.

Conclusion

[7.1] The word fulcrum is defined by The Oxford Pocket Dictionary of Current English as “the point on which a lever rests or is supported and on which it pivots.” Similarly, the point of arrival in the WB Fanfare acts as a pivotal event in the opening title sequence, sending the music off in a new direction in preparation for the arrival of the main title. With this corpus study, we have demonstrated that Steiner uses this fulcrum not only as an opportunity for harmonic, melodic, and temporal variety, but also to communicate the overall tone of the film, which typically correlates with the film’s genre.

[7.2] The corpus study finds that Steiner resolved the WB Fanfare in a remarkable variety of ways—at least 60% of the resolutions in the corpus are unique. Steiner’s resolutions feature many different harmonic and melodic paths and incorporate various combinations of added dissonance, pedal tones, and inversion, resulting in the 57 unique fanfare resolutions listed in Appendix B. By pursuing these alternative resolutions, Steiner reinterpreted what could have been a restrictive factor (the repetitive sameness of the fanfare) as an opportunity for exploration and invention.

[7.3] The corpus study also finds that Steiner used the fanfare’s resolution as a tool to quickly establish the tone of the film. We were able to calculate this tone by using just two pieces of information: the sonority type and melodic resolution at the point of arrival. Our findings illustrate good correlation between film genres and the calculated tone values at the moment of resolution. This resolution (along with the music that follows) gives the audience a clear and concise clue as to what they might expect from the film. As Roy Prendergast notes, the subtle alterations that Steiner uses “can become very valuable if the scene itself is emotionally neutral; the music can add something not already present on the screen” (1977, 220). In the case of the opening title sequence (particularly for Warner Bros. films of this era), the music is the principal actor, providing all of the emotional context for the film, since there is rarely anything happening on the screen except static title cards and rolling cast lists.

[7.4] Following the fulcrum of the fanfare’s resolution, there is a strong connection between visual cues and musical structure in the opening title sequence. The corpus study found that although the transitions were frequently unstable, both tonally and metrically, their principal function was to fill in the space between the fanfare and the arrival of the main title on the screen. James D’Arc notes that Steiner “possessed a unique facility for ‘catching’ things musically, an uncanny ability to match, with split second timing, exact and appropriate sounds with the action on the screen” (2002, 3–4). With this study we have found that Steiner’s desire to synchronize music and imagery extends to the opening title sequences as well.

[7.5] Hugo Friedhofer—Steiner’s orchestrator for many of the films used in this corpus (and an Oscar-winning film composer in his own right)—once stated that “real film music began with Max. Many of the techniques were invented by him, and many of his devices have become common practice. His is true mood music . . . that is also connective tissue, subtle and sensitive” (Behlmer 1973). In this corpus study, we have examined some of this “connective tissue” in Steiner’s music—pivotal events that subtly shift the tone, sensitively provide emotional cues, and carefully connect what we hear with what we see on the screen.

Appendix A. Films Used in the Corpus Study

    Open Appendix A (PDF file)

Appendix B. WB Fanfare Reductions (1937–1955)

    Open Appendix B (PDF file)

Appendix C. Lack of Tonal, Temporal and Metrical Directionality in the Transition

[8.1] We initially hypothesized that the purpose of Steiner’s transitions would be to enable a convincing musical shift from the fanfare to the main title. To determine whether or not this was true, we examined each transition in terms of tempo, key, and meter to see if it (1) moved the tempo closer to the tempo of the main theme, (2) moved the key closer to the key of the main theme, or (3) moved the meter closer to the meter of the main theme. In other words, we wanted to see whether there was temporal, tonal, or metrical directionality from the fanfare through the transition to the main theme.⁽⁷¹⁾

[8.2] Overall, as Example 33 illustrates, transitions moved the music closer to the tempo of the main theme only 48% of the time,⁽⁷²⁾ closer to the key of the main theme only 33% of the time,⁽⁷³⁾ and closer to the meter of the main theme only 10% of the time.⁽⁷⁴⁾

[8.3] Tempo: The tempi established by transitions approached the tempo of the main theme about half of the time (48%). In a handful of cases, significant ritardandi near the end of the transition occurred as well, nicely setting up a new tempo for the main theme. But just as often, contrasting tempi are simply juxtaposed. For example, in Pursued (1947), the fanfare starts at 88 bpm, the transition increases to 138 bpm, and the main theme shifts to 96 bpm. Similarly, in Dive Bomber (1941), the fanfare begins at 84 bpm, the transition decreases to 72 bpm, and the main theme shifts suddenly to 120 bpm. These do not communicate temporal directionality.

[8.4] Key: Just one-third of the examples (22 of 67) featured a transition that “moved closer” to the key of the main theme tonally. Of these, eight moved directly from C major to the key of the main theme, twelve moved to a key that was more closely related to the main theme, and two approached the key of the main theme through a process of linear chromaticism. 13 of the transitions (21%) moved farther away from the key of the main theme. More significantly, 29 of the transitions (43%) were tonally unstable. A few of these unstable transitions clearly arrive on the dominant of the main theme.⁽⁷⁵⁾ But more often, the keys are simply juxtaposed. For example, the transition for Santa Fe Trail (1940) connects the fanfare and main theme (both in C major) with a passage in F♯ major. And in Dallas (1950), Steiner ends an unstable transition with a dominant in E, but brings in the main theme unexpectedly in the key of C. Overall, Steiner’s transitions do not consistently move closer to the key of the main theme, although their very tonal instability might help to set the main theme in better relief when it arrives.

[8.5] Meter: Only 10% of the examples (7 cases) featured a transition that “prepared the meter” of the main theme. This much lower percentage partially reflects that fact that Steiner frequently maintained ${}_4^4$ meter throughout the fanfare, transition, and main theme (in 26 cases, or 39% of the examples). However, in the remaining 34 cases (where the meter does change) the transition clearly does not set up the meter of the main theme. Nine of these transitions are also metrically unstable. For example, in The Voice of the Turtle (1947), the transition shifts between ${}_8^{12}$, ${}_4^4$, and ${}_4^6$ meter, followed by a main theme in ${}_8^6$ meter.

Brent Yorgason
School of Music
HFAC E-566
Brigham Young University
Provo, UT 84602
brent_yorgason@byu.edu

Jeff Lyon
HBLL 6740
Brigham Young University
Provo, UT 84602
jeff_lyon@byu.edu

Works Cited

Annyas, Christian. 2018. “Warner Bros. Logo Design Evolution.” <http://annyas.com/screenshots/warner-bros-logo/>

Behlmer, Rudy. 1973. Warner Brothers—50 Years of Film Music. Liner notes. Warner Bros.

—————. 1985. Inside Warner Bros. (1935–1951). Viking Press.

Biamonte, Nicole. 2014. “Formal Functions of Metric Dissonance in Rock Music.” Music Theory Online 20 (2). <http://www.mtosmt.org/issues/mto.14.20.2/mto.14.20.2.biamonte.html>

Buhler, James, and David Neumeyer. 2014. “Music and the Ontology of the Sound Film: The Classical Hollywood System.” In The Oxford Handbook of Film Music Studies, ed. David Neumeyer, 17–43. Oxford University Press.

Childs, Adrian P. 1998. “Moving Beyond Neo-Riemannian Triads: Exploring a Transformational Model for Seventh Chords.” Journal of Music Theory 42 (2).

D’Arc, James. 2002. Max Steiner: The RKO Years, 1929–1936. Liner notes. Brigham Young University Film Music Archives.

De Clercq, Trevor and David Temperley. 2011. “A Corpus Analysis of Rock Harmony.” Popular Music 30 (1): 47–70.

Duinker, Ben and Hubert Léveillé Gauvin. 2017. “Changing Content in Flagship Music Theory Journals, 1979–2014.” Music Theory Online 23 (4). <http://mtosmt.org/issues/mto.17.23.4/mto.17.23.4.duinker.html>

Easley, David B. 2015. “Riff Schemes, Form, and the Genre of Early American Hardcore Punk (1978–83).” <http://www.mtosmt.org/issues/mto.15.21.1/mto.15.21.1.easley.html>

Gollin, Edward. 1998. “Some Aspects of Three-Dimensional Tonnetze.” Journal of Music Theory 42 (2): 195–206.

Gorbman, Claudia. 1987. Unheard Melodies: Narrative Film Music. Indiana University Press.

Hook, Julian. 2007. “Cross-Type Transformations and the Path Consistency Condition.” Music Theory Spectrum 29 (1): 1–40.

Lehman, Frank. 2012. “Reading Tonality Through Film: Transformational Hermeneutics and the Music of Hollywood.” PhD diss., Harvard University.

—————. 2013. “Hollywood Cadences: Music and the Structure of Cinematic Expectation.” Music Theory Online 19 (4). <http://www.mtosmt.org/issues/mto.13.19.4/mto.13.19.4.lehman.html>

—————. 2018. Hollywood Harmony: Musical Wonder and the Sound of Cinema. Oxford University Press.

Lyon, Jeff, and Brent Yorgason. 2020. “Cataloguing Max Steiner: A Corpus Study of Film Scores.” Journal of Film Music 9 (1–2).

Max Steiner Collection. 1981. MSS 1547. L. Tom Perry Special Collections. Harold B. Lee Library. Brigham Young University.

Murphy, Scott. 2006. “The Major Tritone Progression in Recent Hollywood Science Fiction Films.” Music Theory Online 12 (2). <https://mtosmt.org/issues/mto.06.12.2/mto.06.12.2.murphy.html>

—————. 2014. “Transformational Theory and the Analysis of Film Music.” In The Oxford Handbook of Film Music Studies, ed. David Neumeyer, 471–99. Oxford University Press.

Prendergast, Roy M. 1977. A Neglected Art: A Critical Study of Music in Films. New York University Press.

Richards, Mark. 2016. “Film Music Themes: Analysis and Corpus Study.” Music Theory Online 22 (1). <http://www.mtosmt.org/issues/mto.16.22.1/mto.16.22.1.richards.html>

Rockwell, Joti. 2009. “Birdcage Flights: A Perspective on Inter-Cardinality Voice Leading.” Music Theory Online 15 (5). <http://www.mtosmt.org/issues/mto.09.15.5/mto.09.15.5.rockwell.html>

White, Christopher Wm. 2018. “Feedback and Feedforward Models of Musical Key.” Music Theory Online 24 (2). <http://mtosmt.org/issues/mto.18.24.2/mto.18.24.2.white.html>

White, Christopher William, and Ian Quinn. 2016. “The Yale-Classical Archives Corpus.” Empirical Musicology Review 11 (1).

Yorgason, Brent, and Jeff Lyon. 2017. “The Film Score Themes of Max Steiner: A Corpus Study.” Poster session presented at the Music Informatics Interest Group Meeting for the Society for Music Theory Conference, Arlington, VA.

—————. 2020. “Fanning Out from the Fanfare: Max Steiner’s Theme for Warner Brothers.” Journal of Film Music 9 (1–2).

Footnotes

1. This shield replaced a series of changing designs that included imagery of the Warner Bros. film studios, a Vitaphone banner, and a shield that flew through the clouds (Annyas 2018). This newer shield design remained in use until 1967 (though widened a bit in 1948). Following an exploratory period, the classic WB shield design returned in 1985 and continues to be used today (in various decorative forms).
Return to text

2. For more information about Newman’s fanfare, see Frank Lehman’s discussion in “Hollywood Cadences” (2013). Lehman notes that the purpose of such studio fanfares was “to arrest the viewer’s attention” ([1.2]) and to “[establish] the sound world of a particular movie . . . the sound world of ‘Classic Hollywood’ with its promise of spectacle, big stars, and narratival coherence.” ([1.4]).
Return to text

3. In Hollywood Harmony (2018), Lehman evocatively characterizes Steiner’s open-ended fanfare as “a musical interrobang, forcefully grabbing attention but purposefully eschewing syntactic closure” (2). Among the many studio logos Lehman discusses, Steiner’s is the only one that transitions directly into the main credits, with the others “safely enclosed within a brief, neatly bound package” (2).
Return to text

4. Claudia Gorbman (1987), one of the first musicologists to discuss Steiner’s music in depth, observes that in addition to grabbing moviegoers’ attention, opening title music “defines the genre [and] sets a general mood . . . setting up expectations of the narrative events to follow” (82).
Return to text

5. All dates of Warner Bros. releases were acquired from the “List of Warner Bros. Films” resource at https://en.wikipedia.org/wiki/List_of_Warner_Bros._films. Tovarich was the last film released in 1937.
Return to text

6. Steiner’s hand-written film sketches and original studio recordings are preserved in the Max Steiner Collection in the Special Collections division of the Harold B. Lee Library at Brigham Young University.
Return to text

7. In subsequent cue sheets, Steiner simply refers to the fanfare as “Signature.” We will refer to it as the WB Fanfare.
Return to text

8. Steiner’s fanfare shares many of the common features of studio logos described in Lehman 2018: “cumulatively expanded orchestration, upward-arching melodies, and a sustained crescendo in volume” with music that is “heavily end weighted” in its goal of “celebratorily proclaiming the company that is responsible for bringing a film to its audience” (2). However, other frequent characteristics of studio logos are not present, such as the use of “a single sustained chord” or “simple cadential progressions” leading to a “final, almost invariably major triad” as part of a “self-contained musical [statement]” (2). Instead, Steiner’s fanfare oscillates between chromatic-mediant related harmonies on its way to a destabilizing VII♯ chord that has no prescribed resolution.
Return to text

9. Morgan met frequently with Steiner in his later years (mid-to-late 1960s) and had wide-ranging conversations with him about his life and career, including the events surrounding his composition of the WB Fanfare. Unfortunately, these conversations were not recorded at the time.
Return to text

10. Howard Jackson and Bernhard Kaun, who were generally assigned to B pictures at Warner Bros., did not adopt it immediately, but did so within the year. The first Howard Jackson-scored film featuring the Steiner fanfare is likely Beloved Brat (April 30, 1938), and the first film by Kaun appears to be Blackwell’s Island (March 1939), but there may well have been others—many of the B films scored by Jackson and Kaun during this period are difficult to find today. Franz Waxman used Steiner’s fanfare a few times during his stint at Warner Bros. (1943–1947) for Old Acquaintance in 1943 and Her Kind of Man in 1946. But his first film at Warner Bros.—Air Force (1943)—used original opening music. Erich Wolfgang Korngold had enough star-power to be an exception to the norm at the Warner Bros. studio (and enough time to create new overtures for each film, since he only accepted one or two film assignments per year). However, his opening music for The Sea Hawk (1940) bears some interesting similarities to Steiner’s fanfare as well.
Return to text

11. In his sketch for Gold Is Where You Find It, Steiner writes to his orchestrator Hugo Friedhofer: “The first two bars are a so called W. B. signature - I used it in ‘Tovarich’ - but may be you can make it . . . louder!!!”
Return to text

12. There is also a small difference in the upper strings near the end of the fanfare, with the final pickup gesture in Gold (bracketed in the example) notated a third lower than in Tovarich. Steiner continues to alter this pickup gesture as needed in subsequent versions of the fanfare.
Return to text

13. Having referred back to this film so frequently over the years, Steiner playfully refers to it as “Gold is where you can stick it!” in his sketches for San Antonio (1945) and Cloak and Dagger (1946).
Return to text

14. These films are Sergeant York in 1941 and Mission to Moscow, This is the Army, and The Battle of Britain in 1943.
Return to text

15. These exceptions are the use of dramatic “Indian music” for the opening of They Died with Their Boots On (1941), Tchaikovsky’s Piano Concerto no. 1 for The Great Lie (about a concert pianist) in 1941, and Gershwin’s own music for the biopic Rhapsody in Blue in 1945.
Return to text

16. Steiner actually supplies the WB Fanfare as an “alternate main title” in his sketch for Treasure of the Sierra Madre, but the studio obviously decided to use his new main title music instead. Other notable films not using the fanfare between 1948 and 1955 are The Adventures of Don Juan (1948), The Fountainhead (1949), Beyond the Forest (1949), and The Caine Mutiny (1954).
Return to text

17. The 1948 WB logo was widened and often superimposed over a variety of decorative backgrounds. The logo continued to evolve until the shield itself was omitted in 1972 (Annyas 2018).
Return to text

18. Steiner does resolve to a dominant for Lady Takes a Sailor in 1949, but also adds melodic dissonance and connects the fanfare directly to the main theme.
Return to text

19. See also Murphy’s (2006) corpus study of science fiction scores.
Return to text

20. Attributing authorship is complicated in film music, as multiple composers, arrangers, and orchestrators can be involved in a single film. For 84 of the 88 films listed here, we fortunately have access to sketches in Steiner’s own hand that show he composed the main title music. For the other four films included in this list—Dark Victory (1939), Four Wives (1939), One More Tomorrow (1946), and On Moonlight Bay (1951)—we have cue sheet evidence that indicates Steiner composed the main title music. Cue sheets are official documents prepared by the studio to keep track of composer attributions for royalty purposes. These same cue sheets helped us to rule out several films where Steiner was the principal composer but likely did not create the main title music. For example, Steiner did most of the underscoring for the musicals Night & Day (1946), Love and Learn (1947), and My Wild Irish Rose (1947), but, based on the cue sheet data, the main titles were probably written by Ray Heindorf, who arranged and orchestrated the musical numbers in those films.
Return to text

21. Three other films that he refers back to in his sketches (Crime School, Mark Twain, and Saratoga Trunk) in turn refer back to Gold, so 85% of the films derive at least indirectly from Gold.
Return to text

22. Eleven of the fanfares were written in keys other than C major (B♭, B, C♯, and A major). Four of the fanfares in B♭ major derive from Tovarich—the others he asks to be transposed “one tone lower” from Gold. Confessions of a Nazi Spy and Dodge City (both 1939) are unusually scored in the keys of C♯ major and B major, respectively. The Sisters (1938) is sketched in B♭ major, but Steiner writes a note to his orchestrator to have it transposed another half-step lower to A major: “Vito! These Main Title[s] to be transposed 1/2 tone lower - Max Steiner (from Mr. Friedhofer’s score) - and please have carefully proofread - Grazias! Signore! Si prega de non sputare!! (as Mussolini would say!)”
Return to text

23. Example 12 summarizes the distribution of tempi in the fanfares, most of which ranged from 69 to 88 bpm (with one quite notable outlier at 132 bpm). This outlier, Jim Thorpe, All-American (1951) will be discussed further below.
Return to text

24. The harmonies shaded in gray (labeled Em, Bm, and F♯m) represent resolutions to half-diminished seventh chords, rather than minor sonorities.
Return to text

25. The resolutions to E and G major are to dominant sevenths rather than to triads (see #41 and #49 in Appendix B). All of the other pink-shaded slices represent resolutions to triads.
Return to text

26. Seventh resolutions are not included in this particular chart.
Return to text

27. In that single instance—a resolution to an A♭ minor triad in The Lion and the Horse (1952)—the B is enharmonically respelled as a C♭. See #53 in Appendix B.
Return to text

28. The challenge of working with non-triadic harmonies is one of the limitations of Neo-Riemannian theory. We have handled these chords as follows: chords based on major or minor triads (such as dominant sevenths) are represented by their root, third, and fifth only. Resolutions to half-diminished sevenths are represented by their third, fifth, and seventh only (forming a minor triad; this approach is described in Childs 1998). Only 9 of the 88 resolutions (10%) fell into this latter category.
Return to text

29. Most of the instances of C major come from resolutions that follow the Gold model, which resolves to a C dominant chord.
Return to text

30. In total, 50 of 88 resolutions (57%) feature a chromatic C–B–B♭ line in either the bass or an inner voice.
Return to text

31. Neo-Riemannian theory has become an essential tool for film music scholars, particularly those who study “New Hollywood” scores. The transformations used here are those described in Lehman 2018, involving combinations of P (parallel), R (relative), L (Leittonwechsel), S (slide), N (near fifth), and F (far fifth) and H (hexatonic) transformations, as well as the T_n operation for transposition. Notably absent from the corpus are any simple N, F, or SLIDE transformations—and there is only one P.
Return to text

32. The L’ transformation allows one of the pitches in a triad to bifurcate, moving by semitone in two opposite directions. Using L’, a minor triad can be transformed into a dominant seventh and a major triad can be transformed into a half-diminished seventh. Hook’s L’ transformation does not allow us to map triads onto minor or major seventh chords, so these resolutions have no transformational path given (#4, 9, 13, 19, 36, 37, 40, 54, and 55 in Appendix B).
Return to text

33. These tonal-triadic progression classes (TTPC) were proposed by Murphy (2014) for the analysis of modern film music. Each follows the formula MxM, where M indicates major (M) or minor (m) and x is the number of semitones between the roots of the two triads involved. Our labeling system here uses Murphy’s K equivalence relation (ignoring the key) but it does not use the R equivalence relation (ignoring the order of the triads), so it differs somewhat from Murphy’s approach.
Return to text

34. In order to evaluate the resolutions in the corpus as potential cadences, we interpreted the B major triad at the end of the fanfare as a dominant that would presumably resolve to E minor (although it never does so). See Lehman’s Example 23 for more specific attributes of this cadence type.
Return to text

35. 46 of the 49 fanfares that resolve cadentially (94%) are CMCRs. None of the traditional cadence types (authentic, half, plagal, etc.) occur at the point of resolution.
Return to text

36. The P-matrix was introduced by Rockwell (2009). The top row gives the number of ascending half and whole steps in the progression, and the bottom row enumerates the descending half and whole steps. (A few examples—such as The Unfaithful at #45 in Appendix B—also include a third column for interval-class 3 motions). Rockwell also identifies nine “types” of parsimonious voice leading (see his Example 4). These include all of the progressions that involve three or fewer semitone motions. Whenever Steiner’s progressions map onto one of Rockwell’s nine types, we have indicated it below the P-matrix.
Return to text

37. Note that in this and similar cases, dissonant elements in the resolution (the smaller notes in parentheses) are not accounted for in the total voice-leading measure. For the sake of consistency, we will treat all of these dissonant elements (embellishing tones, dissonant pedals, ninths and elevenths, etc.) as pitches added to the resolution sonority, rather than pitches arrived at through voice leading.
Return to text

38. The reason for distinguishing between different melodic resolutions will become apparent in the following section.
Return to text

39. If we were to consider aspects of orchestration and temporality in addition to harmonic and melodic resolution, this percentage would surely increase.
Return to text

40. Note that if we only wanted to determine how many ways Steiner resolves the B major triad, we would need to eliminate three resolutions from the total, with the percentage of unique resolutions ranging from 60% (53 out of 88) to 38% (33 out of 88), depending upon the criteria of inclusion. Although there are four harmonically-extended fanfares, incorporating #56 (reading #2 for The Adventures of Mark Twain) would simultaneously eliminate #32 (reading #1), since there are no other films that share this resolution.
Return to text

41. Often these thematic borrowings are based on similarities in genre. Crime School (1938), They Made Me a Criminal (1939), and Flamingo Road (1949) are all crime films that use the same main themes (with slight alterations) and fanfare resolution #4; Confessions of a Nazi Spy (1939) and Cloak and Dagger (1946) are both spy films, sharing the same transitional theme—a distorted version of the Nazi anthem “Deutschland Über Alles”—and fanfare resolution #14; and Each Dawn I Die (1939) and Key Largo (1948) are both films noir, sharing main title music and fanfare resolution #16. Four Daughters (1938) and its sequel Four Wives (1939) fittingly use the same main title music, sharing fanfare resolution #7. Other instances of thematic borrowing that result in the sharing of a resolution are: I Was a Communist for the FBI (1951), which reuses the main theme from The Amazing Dr. Clitterhouse (1938)—fanfare resolution #6; Trouble Along the Way (1953) reuses the main theme from The Old Maid (1939)—fanfare resolution #17; In This Our Life (1942) and Lightning Strikes Twice (1951) reuse the transitional theme from Dust Be My Destiny (1939)—fanfare resolution #18; and The Breaking Point (1950) has a similar transition to A Stolen Life (1946)—fanfare resolution #40.
Return to text

42. Lehman (2013) notes that “CMCRs to minor tonics are possible but much less frequent in film music, since they lack the element of joyous exuberance and the fantastic that is the essential reason for their widespread usage in genre cinema” (footnote 58). We agree that CMCRs to minor tonics do not create a feeling of “wonderment,” but would argue that they can be effective in communicating tragedy and other dark emotions. There are several instances of such tragic resolutions in the fanfare corpus (accounting for 18% of the cadences).
Return to text

43. Steiner used only two other cadence types (the remaining 6%). These resolve to unexpected tonics that are diatonic rather than chromatic (#45 resolves from V to iv, and #53 resolves from V to iii).
Return to text

44. One semitone was shifted in 4 resolutions; two were shifted in 13 resolutions; three were shifted in 26 resolutions; four were shifted in 29; five were shifted in 6; and six were shifted in 9 resolutions.
Return to text

45. Type III involves two upward semitonal shifts. Tabulating the types, we see that Steiner uses Type I 5 times, Type II 10 times, Type III 0 times, Type IV 3 times, Type V 3 times, Type VI 10 times, Type VII 11 times, Type VIII 2 times, and Type IX 2 times. If anything, Steiner seems to prefer voice leading involving Type VI (to B♭ or C) and Type VII (to D, F♯, and A♭). Overall, 46 of Steiner’s resolutions (52%) were given a type, and would thus be considered parsimonious by Rockwell’s measure.
Return to text

46. Although there were no main themes occurring literally in C♯ minor or F♯/G♭ major, these keys did emerge through transposition. On the other hand, two themes were literally in the key of B minor, but when transposing, became C♯ minor.
Return to text

47. Indeed, we are not aware of any pre-existing models for defining the tone of a film. Thus, the values chosen in Example 19 are necessarily subjective to some degree. We welcome critique regarding our choices here and invite readers to consider what their own criteria for evaluating the tone of a film might be. Note that, once determined, these values were not subsequently altered to better suit our hypotheses or to improve the results.
Return to text

48. In most cases, melodic resolution was determined by listening to the lead horn part. In other cases, it was determined that the resolution was provided in the strings.
Return to text

49. The abbreviations in Appendix B may be interpreted as follows: APP = appoggiatura, LT = leading tone, RET = retardation, CT = common tone, SUS = suspension, and DEFLECT = deflection.
Return to text

50. Curiously, Steiner never resolved the fanfare to a fully-diminished sonority.
Return to text

51. The notes in parentheses are the ones that were treated as added dissonances.
Return to text

52. Results can also be seen in Appendix B, with the calculated tone for each fanfare listed at the far right.
Return to text

53. For instance, The Beast with Five Fingers (1946), a horror film about a decapitated hand that comes to life to play Bach and commit mayhem, surely merits a darker tone than -20.
Return to text

54. Film genre labels are notably inconsistent in terms of what they attempt to describe (whether it be the subject matter, the setting, the mood, or the audience). And these labels may not fit any particular film that well—certain films defy generic labeling. Modern genre designations are also problematic in that they may classify Classic Hollywood films in terms that were not then in common usage, such as the term “film noir” for films that were then known simply as crime films or melodramas. Other classic genres, such as the “woman’s film” (a genre for which Steiner wrote many fine film scores) are no longer classified as a separate genre. However, it is not our goal to create a uniform theory of film genres or to critique existing labels. Although imperfect, we have found that the IMDb genre designations are not inaccurate, and have preferred to use an external, established source of information rather than to create our own, potentially biased genre labels, particularly since we are dealing with a large corpus.
Return to text

55. 71% of the films studied (64/90) were tagged “drama” (making it a rather unhelpful descriptor). The next most common genre was “romance” (35/90, or 39%). Given their ubiquity, it is not surprising that these two genres involved a fairly even distribution of resolution types.
Return to text

56. Note that Daughters Courageous (1939) was recorded in B♭. All of the others are in C major.
Return to text

57. Two-thirds of the appoggiatura resolutions, three-fourths of the deflections, and three-fifths of the downward leaps in the corpus overall were tagged “romance.”
Return to text

58. There were no “leap down” resolutions in this genre.
Return to text

59. Although this may seem odd to modern movie-goers accustomed to the more somber, dark, and pessimistic tone of war films like Schindler’s List (1993), Platoon (1986), or Apocalypse Now (1979), propagandistic war films in the 1940s were more often heroic, optimistic, and uplifting in nature. And of all the Hollywood studios, Warner Bros. was the most enthusiastically supportive of government efforts in its use of war propaganda.
Return to text

60. Appendix C provides a summary of our findings regarding this initial hypothesis. Although it did not prove to be accurate, the findings presented here do provide a good summary of Steiner’s use of tonal, metric, and temporal juxtaposition.
Return to text

61. 12 of the 21 examples with no transition (57%) feature strong tonal juxtapositions. For example, Arsenic and Old Lace (1944) moves directly from C major, 4/4, 84 bpm to Db major, 2/4, 120 bpm.
Return to text

62. In other cases, the name of the star of the film coincided with the appearance of the main theme, followed immediately by the title. This technique may serve to emphasize the stars more than the title. See, for instance, the emphasis on Humphrey Bogart and Lauren Bacall’s names in The Big Sleep (1946), Bette Davis’s in A Stolen Life (1946), and Edward G. Robinson’s in A Dispatch from Reuters (1940).
Return to text

63. Note that the opening title sequence does not ordinarily end after the appearance of the main theme. Depending on the length of the title sequence, any number of subordinate themes may occur following the main theme. In our future research, we hope to examine more thoroughly the overall structure of opening title sequences within the Max Steiner corpus.
Return to text

64. The transition may have other functions such as establishing the setting of the film. For example, in the transition of Virginia City (1940), two quick quotations of “Dixie” and “Battle Hymn of the Republic” help establish the Civil War era (with a Confederate soldier as the lead character).
Return to text

65. The most common transition lengths are 8 beats (occurring 15 times), 4 beats (10 times), 12 beats (9 times), and 16 beats (6 times). Thus, Steiner does demonstrate a preference for hypermetric regularity in the transition, choosing patterns that will maintain continuity with the meter of the fanfare.
Return to text

66. In some of these cases, the starring actor is billed with the main title on-screen.
Return to text

67. The fanfare in Dodge City (1939) is similarly extended through repetition to allow its resolution to coincide with the appearance of the main title (see Video Example 26a above). Thus, the role of the extended fanfare may be similar to that of the transition: to delay the arrival of the main theme. Overall, six of the eight extended fanfares lead directly into the main theme, with no separate transition.
Return to text

68. Other fanfare extensions occur in We Are Not Alone (1939), City for Conquest (1940), The Adventures of Mark Twain (1944), Arsenic and Old Lace (1944), and I Was a Communist for the FBI (1951). Another example of harmonic deviation can be found in the fanfare for Mildred Pierce (1945)—although it is not extended, it is the only example that does not end in B major, introducing a dissonant B♭ pedal before moving to A♭m7 over B♭ (see #37 in Appendix B). This additional dissonance might be intended to convey the crashing of waves on the screen—creating the “wave effect” that Steiner indicates in his sketch.
Return to text

69. Other examples involving sound effects at the point of resolution include Dive Bomber (1941), which resolves to a low trill that accompanies rumbling plane engines; The Breaking Point (1950), which resolves to the sound of seagulls on the beach; and Battle Cry (1955), in which the transition begins with lightning and a peal of thunder.
Return to text

70. This toy version of the fanfare resolves to D major with a LEAP UP, using model #10 from Appendix B. This very bright resolution has a calculated tone of 100. This resolution was not included in the corpus.
Return to text

71. Only 67 films were included in this measure, since 21 of the films had no transitional section, with the main theme following the fanfare directly.
Return to text

72. We defined “tempo moving closer” as follows: the difference between the tempo of the transition and the main theme (as measured by number of beats per minute) is smaller than the difference between the tempo of the fanfare and the main theme. For example, in the film Dallas (1950), the tempo of the fanfare is 80 bpm, the tempo of the transition is 104 bpm, and the tempo of the main theme is 120 bpm. The tempo difference between the transition and the main theme is 16 bpm while the tempo difference between the fanfare and the main theme is 40 bpm. Thus, the transition “moved closer” to the tempo of the main theme.
Return to text

73. We used the circle of fifths to define “key moving closer,” as follows: if the key of the main theme is closer to the key of the transition (on the circle of fifths) than to the key of the opening fanfare, then the transition has moved closer to the key of the main theme. For example, in Captains of the Clouds (1942), the fanfare begins in C major (0 flats), the transition is in F major (1 flat), and the main theme is in B♭ major (2 flats). Since the key of the transition is nearer on the circle of fifths to the key of main theme, the transition has “moved closer” to the key of the main theme.
Return to text

74. For meter, we defined “moving closer” as any meter change between the opening ${}_4^4$ meter and the meter of the main theme in which (1) the number of beats in the transition changes to the number of beats in the main theme, or (2) the beat division changes from simple to compound for a compound-meter main theme. For example, in A Kiss in the Dark (1949), the fanfare begins in ${}_4^4$, the transition moves to ${}_4^3$ and the main theme is also in ${}_4^3$. Since the number of beats changed from four to three, the transition “moved closer” to the meter of the main theme.
Return to text

75. Some examples of this are The Bride Came C.O.D. (1941), My Girl Tisa (1948), Deep Valley (1947), On Moonlight Bay (1951), Trouble Along the Way (1953), and Battle Cry (1955).
Return to text

Most of the instances of C major come from resolutions that follow the Gold model, which resolves to a C dominant chord.

Copyright Statement

Copyright © 2020 by the Society for Music Theory. All rights reserved.

[1] Copyrights for individual items published in Music Theory Online (MTO) are held by their authors. Items appearing in MTO may be saved and stored in electronic or paper form, and may be shared among individuals for purposes of scholarly research or discussion, but may not be republished in any form, electronic or print, without prior, written permission from the author(s), and advance notification of the editors of MTO.

[2] Any redistributed form of items published in MTO must include the following information in a form appropriate to the medium in which the items are to appear:

This item appeared in Music Theory Online in [VOLUME #, ISSUE #] on [DAY/MONTH/YEAR]. It was authored by [FULL NAME, EMAIL ADDRESS], with whose written permission it is reprinted here.

[3] Libraries may archive issues of MTO in electronic or paper form for public access so long as each issue is stored in its entirety, and no access fee is charged. Exceptions to these requirements must be approved in writing by the editors of MTO, who will act in accordance with the decisions of the Society for Music Theory.

This document and all portions thereof are protected by U.S. and international copyright laws. Material contained herein may be copied and/or distributed for research purposes only.