Author Archives: Paul Roth

The Dance Bands of Pittsburgh

In 2013, WESTERN PENNSYLVANIA HISTORY MAGAZINE published Paul Roth’s retrospective article/memoir on dance bands.

The article, “PITTSBURGH’S DANCE BAND ERA,” is both a nostalgic reminiscence about the bands of the Western Pennsylvania region, but also about the overall dance band culture.

Click on the image to download the PDF of this article.

The Dance bands of Pittsburgh cover photo

Recording the Alto Bands

Commencing in about 1928, Columbia started releasing records by Guy Lombardo and His Royal Canadians.

This was an innovation, for Lombardo’s style of arrangement – separated horn sections, with the saxophones led by a strong alto-played melody, which I call the “alto” sound – had rarely been heard on any previous recordings. Columbia and Lombardo had many popular issues until they parted company in 1932.

Lombardo signed on with Brunswick, a major competitor. It appears that Columbia – dedicated to the Lombardo style – then started to record Art Kassel, a Midwestern unit with a similar sound. The other major record producer, Victor, had at this time no competing alto band until it signed Kassel in 1934 for the Bluebird label and Jan Garber in 1933 for the Victor label. At this point all three major labels were committed to propagate the alto sound. Kassel remained with Bluebird until 1942.

Then, in 1934, Lombardo jumped to the newly-formed Decca label.

This left Brunswick without an alto band. They started recording Kay Kyser in 1935. In 1936, Lombardo and Garber changed sides, with Garber going to Decca and Lombardo to Victor.

In 1937, Vocalion – Brunswick’s “budget label”- released Sammy Kaye’s first offerings. And Bluebird fielded an array of territory bands, such as Jimmy Livingston and his Hotel Charlotte Orchestra, which made a dozen sides in 1937-38.

In 1938, Garber left Decca for Brunswick and Lombardo returned to Decca. a relationship which would last well into the 1950s. In the same year, Kaye, who had several hits on Vocalion, jumped to Victor and, like Lombardo, stayed into the 1950s. And what about Bluebird? They recorded Blue Barron from 1938-through 1940,

In 1939, with Brunswick defunct, Kyser switched to Columbia, which also absorbed the Vocalion label, with its alto bands, Ray Herbeck and Dick Barrie. Kyser’s recording for Columbia was continuous until he retired about 1951.

Thus, by 1939, things had stabilized between the alto bands and the major labels.

In 1940, disruption occurred again when Garber, Barron, and Kaye signed on with the new U.S. Record Co. (Varsity, Elite, Hit, etc. labels) Kaye, however, released about a dozen sides and then returned to Victor.

By 1942, the war had disrupted the bands, and the recording industry was greatly affected by the long recording ban imposed by the Musician’s Union. At war’s end all of the major alto bands were able to record – some on new labels such as M-G-M and Mercury – but only Lombardo and Kaye could sustain through the 1950s. Again, Lombardo and Kaye were popular enough to foster “ghost” bands which were still playing in the late 20th Century.

It should be noted that this straightforward way of playing the melody sustained these units as “dance” bands long after the swing organizations had left the ballroom.

Copyright 2014, Paul F. Roth. All rights reserved.

Style “Gimmicks” In Dance Bands

Listening to the dance bands of the period 1929 through, say, 1935, it becomes apparent to me why there was an increasing amount of arrangement-style gimmickry applied for several years, commencing in 1936. With a few exceptions, they all sounded the same!

Music recorded during this period was primarily for dancing. Tin Pan Alley’s mass product was songs of 32 bars, meant to be danced to. The arrangers also appeared to have established a common standard: let’s call it the “Leo Reisman” style.

What was this style?

Instrumentally, the melody was carried by the string section, by the brass section, or by the piano. Touches were added by solo tenor sax, clarinet, or muted solo trumpet. Rarely was the saxophone section or brass ensemble given the lead to play.

Listen to the aforementioned Leo Reisman, also Eddy Duchin, Richard Himber, Henry King, Lud Gluskin, and others. This rather conservative style was meant for the hotel-room setting, not the ballroom. The ballroom give birth to a different, less subtle style of arrangement.

There were exceptions. Guy Lombardo, in 1926, followed by Jan Garber in 1933 created what I call the “alto” sound: a 3-or-4 man sax section led by an alto sax with ample vibrato. Hal Kemp, in the early 30s created a variation of the Lombardo style, using muted clarinet ensembles. Both made heavy use of muted, staccato brass.

For the ballroom sound, around 1930, the hot bands began to play ballads with the same string less voicing as Lombardo et al., but usually with the alto-led sax section replaced with a balanced alto/tenor combination. And the brass were more open.

So, in about 1936, the bands – as far as recordings and radio performance were concerned – played ballads in three styles of arrangements, with most playing in “hotel” style.

As radio remote broadcasting became more important in developing a band’s popularity, there commenced in 1933 a
trend to add features which distinguished one band from another.

This was done by adding sound “gimmicks.”

In the hotel bands, Freddy Martin established instant recognition with his smoothly-played, prominent tenor sax tone. Then Ted FioRito added temple blocks and clipped brass choruses. But the most distinguishable sound was imparted in 1936 by Shep Fields with his bubble sound accompanied by glissando solo viola plus the ‘clip-clop’ accents. Later Russ Morgan adapted his “wah-wah” trombone sound to distinguish his band. And Lawrence Welk put into the foreground accordion and clarinet ensembles.

But the most variety in gimmickry was experienced by the alto bands, which started to proliferate about 1933.

Ted Weems had a whistler do a chorus on specialty numbers.

Kay Kyser made regular use of singing song titles (originated in the late ‘20s by several bands including Weems) and the announcement of the vocalist over a theme song vamp after the first (instrumental) chorus. He also borrowed Kemp’s triple-tonguing trumpets.

Sammy Kaye, who started to record in 1937, used the same stylistic features as Kyser and later added glissando accents by an electric steel guitar or alternately by solo trombone. He was closely copied by Blue Barron. Kaye told me that there was a practical side to the singing song titles and announcing the vocalist during the song: elimination of the need for an announcer on-site during a radio remote broadcast.

The ultimately-gimmicked alto band was Horace Heidt’s which employed, at times, all of the Kyser-Kaye features. Plus the addition of a singing choir and a whistler.

Of course, in 1936, there began the age of the swing bands which eschewed such ‘unhip’ gimmickry when playing ballads. Yet one – Glenn Miller’s – achieved a measure of instant recognition with its clarinet-over sax section ballad sound.

This variety of arrangements and sound continued until WWII, after which all conventions were uprooted, especially the apparent need for gimmicks in light of changing musical tastes and the loss of radio appearances, although Kaye maintained his stylistic effects for occasional use well into the ‘70s.

But for a while, it made for an interesting sideshow.

Copyright 2014, Paul F. Roth. All rights reserved.

The Popular Song and Entropy

In Physics, one of the Laws of Thermodynamics theorizes that as complex systems age, they exhibit increasing “entropy”, that is, their elements become more and more disorganized. For instance, as a gas cools, the paths of its ions exhibit increasingly random and unpredictable behavior.

The popular song appears to be exhibiting entropic tendencies.

The American popular song was invented for singing, then for dancing. During its evolution from Stephen Foster’s time to the present it has become so disorganized that the concept of a “song” has devolved into a collection of noises – rhythmic, but disorganized.

Take the “standard” 32-bar pop song, promulgated from the 1920s to provide music for social dancing: fox-trot, waltz, and other steps. It traditionally followed a rigid format of AABA or ABAB, each letter representing an 8-bar segment. A 32-bar song conveniently fit into the 3 minutes provided on a 78-rpm record, with 2 to 4 choruses played, depending on tempo.

The performance of this music was by musicians and singers who read the music score and maintained the appropriate time, provided by a leader or conductor. Groups of musicians varied from a soloist to dance bands of up to 20 personnel. By the 1920s the larger units had grouped into ”sections” of similar instruments. The elements of an arrangement were melody, chords, rhythm, and lyrics.

As related by Artie Shaw and others, many instrumentalists and vocalists. tired of the commercial approach to music – according to the notes – and began to improvise or “jazz-up” the music. At first they played around the written melody. As time progressed, starting in the late ‘40s, they dropped the melody completely and improvised around the chords. “Bop” music was created in this way. Vocal improvisation frequently ignored the actual lyrics. In each case – vocal and/or instrumental soloists – the greater the level of improvisation, the “hipper” or “cooler” the musician was considered. ”Rock” music – heavily employing amplified string instruments – further de-emphasized the ensemble. And since each group tended to write their own songs, in time the roles of composer and musician became merged.

Further disorganization was facilitated by the introduction of the long-playing record, tape recording, and the CD – each of which allowed a composer, musician or singer to completely abandon the structure of a traditional song, or to improvise in “free-form”, frequently with no reference to chord-structure and little time constraint.

Also the amalgamation of the traditional pop song with elements of ethnic, country, soul and other music further induced variation to the formula.

Thus, in a little over 50 years, the generic popular “song” has disorganized into a sort of cacophony through the partial loss of three organizing elements: cohesive melody, chord structure and lyrics – thus exhibiting a sort of entropy. What is left is typically rhythm-based (or in Harold Arlen’s words: “percussive”) or somewhat atonal music.

It may be noted that classical music – to an extent – has undergone entropy, at least in melodic and rhythmic structure: “modern” music seems to be atonal and arrhythmic.

This subject brings to mind a Sid Caesar TV sketch where the bandleader, played by Caesar in a “bopper” beret and glasses, introduced one member of his combo as “playing radar – to tell us when we get too close to the melody.”

Copyright 2014, Paul F. Roth. All rights reserved.