I've always loved Hector Berlioz's Harold en Italie ("Harold in Italy"), even though critics tend to slight it compared to his other ground-breaking symphonies – the Symphonie Fantastique that preceded it and Romeo et Juliet that would follow. Perhaps it's the title – both my late father and older son are named Harold.

The tale of the creation of Harold in Italy has often been told, yet it remains a fascinating confluence of celebrity, musical roots, literary inspiration and, above all, the composer's own quixotic personality. But first a strong caution – the primary source of our information is Berlioz's own fanciful Mémoires. Written largely in 1848, serialized in Le Monde illustré in 1858-9, and published in 1865 for limited private distribution, the Mémoires consisted of travel journals recounted well after the fact, critical essays and biographical glimpses into his life and work. Berlioz issued it as a book at his own expense at the end of his life, largely to defend his reputation and craft an affirmative legacy. While its criticism is ardent, sharp and unfettered ( "It is now time to have done with all this acclamation for Mozart, whose operas are all alike and whose cool beauty is tiresome and distressing"), Pierre Citron notes that biographers have found extensive inaccuracies ranging from honest errors of recall and blurring of detail to outright fabrication. So while the Mémoires serve as a fully accurate and compelling portrait of Berlioz's memories, self-image and convictions, the facts he purports to recite may not be wholly reliable.

With that in mind, Berlioz wrote that after a concert of his Symphonie Fantastique "a man with long hair, piercing eyes and a strange and haggard face" introduced himself.

Paganini in 1831

It was Niccoló Paganini, the cult performer of his era, hysterically worshipped by public and critics alike for his phenomenal virtuosity. Serious composers and even rivals respected his unprecedented skills and novel techniques for revealing expanded musical resources. In January 1834 Paganini asked Berlioz to write a concerto to introduce a new Stradivarius viola he had acquired.

Why in the world did Paganini seek Berlioz for this commission? Their relationship was strained – months earlier, Paganini had declined to play at a benefit Berlioz had organized for Harriet Smithson, an actress upon whom he doted and later married, while Berlioz was conspicuously absent from a gala Paganini recital. Indeed, while surely aware of his reputation, Berlioz never heard Paganini play. And why for a solo display piece? Berlioz was known for his restless orchestration, shifting textures and formal experimentation. Perhaps, to his lasting credit, Paganini recognized a fellow visionary who could lift his artistry to new heights. In any event, they announced to the press a mammoth work for orchestra, chorus and solo viola to be entitled Les dernier instants de Mary Stuart ("The Last Moments of Mary Stuart"), the "Queen of Scots" who had been executed in 1587 after two decades of imprisonment and political intrigue.

According to Berlioz, he tried to combine the solo lines with the orchestra, feeling sure that Paganini's incomparable execution would enable him to give the solo instrument all its due prominence. Yet, when presented with the first movement, Paganini rejected it as having too many rests, insisting that he wanted to be playing all the time. Berlioz then recast the work as a series of orchestral scenes "in which the viola finds itself mixed up [while] always preserving his individuality."

But Paganini's connection with Harold did not end with his disavowal of the work. After attending a December 1838 performance, sapped by the illness that already had taken his voice and before long would take his life, Paganini dragged the composer back on stage, knelt down and kissed his hand. The next day, Paganini's son delivered an ecstatically flattering letter in florid Italian that began: "Beethoven spento non c'era che Berlioz che potesse farlo rivivere." ("Beethoven is dead and Berlioz alone can revive him.") Included was a draft for 20,000 francs (about twice Berlioz's annual earnings) to be presented to Baron de Rothschild. (Many have speculated that the funds were from an anonymous admirer – possibly his publisher Armand Bertini – since Paganini was in financial straits after a disastrous investment in a Paris casino. Ernest Newman, though, feels that Paganini hoped to benefit from the publicity surrounding the gesture to counter widespread criticism of his stinginess.)

Berlioz portrait by Signol

When Berlioz went to give thanks, Paganini said that hearing Harold was the greatest pleasure of his life: "You will never know how your music affected me. It is many years since I had felt anything like it." In gratitude, Berlioz embarked on the composition of a work worthy to be dedicated to Paganini – his astounding Romeo and Juliet (which Paganini would not live to hear).

Harold's musical roots are equally intriguing. After three unsuccessful attempts, in 1830 Berlioz had finally won the prestigious Prix de Rome, a four year scholarship awarded by the French government for study in Italy, which required annual compositions. Berlioz fulfilled his obligation for 1831 with his Rob Roy Overture, for which critical opinion has been sharply divided. Francis Tovey calls it "quite an engaging work," while Hugh Macdonald dubs it "vacuous and repetitive," and suggests that Berlioz deliberately wrote it with the assumption that official taste of the conservative Academy preferred the commonplace to the inspired. Berlioz himself called it "long and diffuse," and claimed to have been so disappointed with its reception at its first and only performance that he burned the score (although the copy he submitted to the French Academy survived). Yet, Berlioz clearly retained some affection for it, as its two primary themes, as well as a lengthy development section, play prominent roles in Harold. Its full title – Intrada de Rob Roy MacGregor – explains much of the Scottish flavor that carries over into Harold – as does the initial concept of the Mary Stuart piece. Indeed, D. Kern Holomon notes a striking similarity of a key Harold theme to the Scotttish song "Scotts wha' hae wi' Wallace bled."

Although traditional in comparison to the five-movement Symphonie Fantastique (and the seven-movement vocal Romeo et Juliet), Harold shares with the former work an idée fixe

The viola's idée fixe of Harold in Italy

– a main theme that pervades the entire work, both directly and, as David Cairns has noted, more subtly as the generator of most of the other tunes, which spring from its component phrases and rhythms. Yet, Cairns observes, unlike in the Fantastique, the Harold idée fixe doesn't transform its character but remains a fixed point of reference for the changing scenes through which the hero passes, coloring them with his poetic awareness, exuberance, introspection and anxiety.

As for literary inspiration, Berlioz claimed that his new work was written in the style of Lord Byron's immensely popular 1812-18 epic poem Childe Harold's Pilgrimage. Berlioz must have closely identified with Byron's title character, a melancholy dreamer who visits and comments upon sites of classical antiquity in search of meaning to counter his own world-weary disillusionment. Although Berlioz desperately had sought the Prix de Rome, once he got to Rome he wrote in his Mémoires that his life there was "a continual martyrdom – one's beautiful musical dreams are dispelled by grim and hopeless reality; every day brings fresh disappointment – while other arts flourish, displaying their manifold beauties, music alone is degraded to the level of a poor hunted slave." His boredom soon turned to wanderlust, as he fled his residency to wander the Italian countryside, gathering impressions, dreams and inpirations that would infuse his new work. Even so, beyond a subconscious autobiographical affinity there is no direct connection between any specific elements of the Byron poem and Harold; Olin Downes quips that the work really should be called "Berlioz in Italy."

Berlioz may also have been attracted to the Byron poem as an extraordinary technical feat. Despite its extreme length, each stanza is written in a strict form consisting of eight lines of iambic pentameter and a final line of iambic hexameter in a rhyme scheme of ababbcbcc. Berlioz's Harold, too, is an uneasy alliance of classical technique and free-thinking attitude. Indeed, despite its lesser status in the Berlioz symphonic canon, Harold displays all of his most important hallmarks.

Many historians regard the most significant formal innovation in Berlioz's three great symphonies (I'm purposely excluding his turgid, ceremonial 1840 Grande Symphonie Funèbre et Triomphale) to have ushered in an era of program music. Paul Henry Lang notes: "Berlioz raised program music from a rather occasional indulgence to a constructive principle of composition." Yet, Lang notes, it spawned inherent friction, as the descriptive elements of Berlioz's literary plan were the antithesis of symphonic abstraction and deflected the natural flow of the music. To Lang, rather than let his unfettered emotion run free with spontaneous feeling, Berlioz tended to lapse into outmoded formal clichés. J. H. Elliott summed it up: "Berlioz's best is wonderful, his worst appalling – and the twain, with the degrees between them, are inextricably confused together." Or, in Harold Schonberg's more flattering assessment, all Berlioz's work is a mixture of flaws and genius; moments of inspiration alternate with banalities, overwriting, self-conscious posing, weak melodies and awkward transitions, yet all such blemishes wither before his prodigious power, originality and ardent Romanticism.

In furtherance of the program, each movement of Harold begins with a descriptive title that tends to draw more attention than the formal structure. Egon Kenton posits that the titles were grafted on as an afterthought, to ease the audience's shock of listening to novel music they might otherwise not be able to grasp. The first is entitled: "Harold aux Montagnes. Scènes de mélancolie, de bonheur et de joie" ("Harold in the mountains. Scenes of melancholy, happiness and joy"). While often described as a modified sonata form, over half its length is a slow introduction launched with a sinuous fugue, as if to suggest the academic doldrums from which Berlioz sought escape in nature. Only after the languor is dispelled by the Harold theme does the sonata form emerge and become apparent.

While commentators may (and often do) debate the success of his formal merits, there can be no dispute as to two technical areas in which Berlioz transformed music. The first is orchestration. In his authoritative History of Orchestration, Adam Carse dubs Berlioz "by far the most progressive, original, independent and daring orchestrator of his time." Indeed, Carse suspects that Berlioz sometimes built up music in order to show off a preconceived instrumental effect (and perhaps, in that process, he generated the stretches that in a traditional sense seemed uninspired). Yet, Carse notes with irony that Berlioz's music never became popular enough to significantly influence either his contemporaries or successors to adopt his innovations. (Julian Rushton suggests that Berlioz relished the conflict between audiences' generic expectations and his musical reality.)

More than any other factor, Berlioz's daring instrumentation arose from seeming deficits in his training – after Haydn, he was only the second great composer who had not risen from the ranks of virtuoso performers. He also had little interest in most music of the past.

Berlioz portrait by Courbet

As a result, he was able to investigate the properties of each instrument without preconception so as to grasp their intrinsic properties, suggest fresh techniques (woodwind mutes, string harmonics, bowing textures), imagine new combinations to realize his expressive purposes, and demand players with the expertise to realize the effects he envisioned.

Berlioz poured his extensive yet intuitive knowledge of instrumentation into one of the classics of music literature, his 1843 Grande Traité d'instrumentation et d'orchestration modernes which explores with great thoroughness the impact of each instrument on listener perception and remains relevant. For Joel-Marie Fouquet, through this treatise Berlioz evolved the science of instrumentation into the art of orchestration. The second movement of Harold is an extraordinary display of Berlioz's skill in that realm. Its title: "Marche de pélerins chantant la prière du soir" ("March of pilgrims singing the evening prayer"). The first section comprises 16 repetitions of a gentle pilgrim's march theme over a walking bass that sustains interest through subtle variation, both of the theme itself and the timbre as it wends its way through various instrumental combinations. After a central section of a religious canto that builds in an urgent harmonic progression and an abbreviated repeat of the opening comes a remarkable coda of breathless tension in which sustained clashing notes of b (horns) and c (harp, oboe and flute) alternate 11 times before relaxing into a concluding E-major chord. Ironically, although this movement was the most popular segment of Harold at the time (often performed alone) and thus was considered the most conservative, its slightly altered persistent repetitions now can be seen as a thoroughly modern harbinger of minimalism.

Berlioz's other undisputed realm of mastery is rhythm. Théophile Gautier likens his rhythmic sophistication to that of poet Victor Hugo, who disdained the simple lines of classical art and used devices to vary the monotony of poetic phrasing. Indeed, Harold has an, edgy, natural feel that defies bar lines and strict timing with syncopation, dropped beats and unexpected accents. Once we know the piece, following the score can be both frustrating and exhilarating, as we can feel a great tension as Berlioz forced his free-wheeling conception into the rigid conventions of notation.

The third movement of Harold proudly displays Berlioz's rhythmic prowess. Entitled "Sérénade d'un Montagnard des Abruzzes à son maîtresse" ("Serenade of an Abruzzian mountaineer to his sweetheart"), it opens with a jaunty rustic oboe and piccolo theme over a sustained open fifth drone and a peppy, syncopated string rhythm, then lapses into a slow, plaintive English horn melody over gently shifting string harmonies. The opening rhythm, forlorn melody and a lazily augmented version of the Harold theme return in an astounding coda where, remarkably, they are overlaid as three independent events occupying the same sonic space, a thoroughly baffling complexity in the context of its era yet a harbinger of the autonomous events of 20th century "chance" music.

Even beyond their evocative titles, each of the first three movements are enriched by resonances of the composer's travels.

J.M. Turner: Childe Harold's Pilrimage - Italy (1832), now in the Tate Gallery, London

In chapter 37 of his Mémoires he spoke of how he spent all the time he could in the mountains of Italy, "free to ignore the flight of time, to despise ambition, to laugh at glory … to wander at will, to dream, to lie enchanted and panting whole days in the soft, hot, murmuring sirocco wind. Perfectly, utterly, illimitably free." The unabashed joy of the first movement clearly stems from such ardent affection. He also wrote of "returning gleaners from fields singing soft litanies to the accompaniment of the sad tinkling of the distant convent bell" – a clear description of the march and coda of the second movement – and, while generally condemning modern Italian music, he admired Palestrina and the reverie produced by calm harmonies without melody or rhythm during a Sistine Chapel service – the generator of the middle section. In chapter 38, he described a leather-lunged serenader accompanied by mandolins and bagpipes, and in chapter 39 the "long series of gay and refreshing refrains repeated over and over" by wandering piffero (fife) players who descend from the mountains – all evident in the elements of the third movement.

While much of Harold is grounded in the composer's cherished memories, the fourth movement culminates the work with a flight of pure fantasy that deliriously displays all the hallmarks of Berlioz's style. It begins, though, with "Souvenirs des scènes précédentes" ("Memories of the previous scenes") as the solo viola recalls to the full orchestra the themes of each prior movement – a conscious throwback to the finale of Beethoven's Ninth Symphony. (Actually, the paradigm's scheme is reversed – Beethoven's orchestra offers prior themes which angry celli rebuff and then emerge with their own soothing "Ode to Joy," whereas Berlioz's ensemble sympathetically supports the viola's citation of earlier ideas and melds them into its own upbeat fervor.)

Then begins the true finale – "Orgy des brigands" ("Brigands' orgy"). Jacques Barzun notes that brigands were rebels who turned to nature to heal the wounds of society, and that Berlioz seized upon their revelry as a violent purging that symbolized an antidote to the repressions of conventional life. In his Mémoires Berlioz describes the scene as "a furious orgy where wine, blood, joy, rage, all combined, parade their intoxication … the brass seem to vomit forth curses and to answer prayers with blasphemies; where they laugh, drink, fight, destroy, slay, violate and utterly run riot." Yet, Berlioz knew that ten minutes of relentless bombastic din would be ineffective, and so added a complementary vision that: "violins, basses, trombones, drums and cymbals all sang and bounded and roared with diabolical order and concord." Thus, the fury is balanced by quiet moments and even the mighty climaxes shimmer with what Carse salutes as shrewd orchestration and a meticulous tonal balance. Indeed, Berlioz prefaces the frenzied culmination with a brilliant contrasting touch – a soft off-stage string trio that wistfully recalls the peaceful pilgrim's march before the revelry overrides it. Rhythmic complexity abounds, bar lines virtually disappear, duple and triple meters are overlaid, and, in a novel touch, several emphatic held notes are prefaced with a tied sixteenth that disrupts the expected downbeat with a sense of great anticipatory urgency.

Perhaps the greatest challenge Berlioz set for himself in Harold was to develop the long-neglected viola into a featured voice. String historian Tully Potter considers the instrument inherently unstabile and treacherous to play at both ends of its range, which lies a fifth below the brilliance of the violin yet an octave above the richness of a cello, and notes that in order to obtain acoustical balance it would have to be too large to play as a shoulder instrument; thus, its compromised size imparts a nasal, throaty tone to its middle register. While they routinely included violas in orchestra string sections, composers of the Romantic era displayed none of the interest in the viola as a solo instrument that had enriched music of the 18th century and would revive in the 20th.

Lionel Tertis (1876 - 1975) with his viola

Indeed, in his Mémoires Berlioz deplored the "antique, absurd and deplorable prejudice that had handed over the performance of the viola parts to second or third-rate violinists. Whenever a violinist is mediocre, it is said, 'He would make a capital violist.'"

Berlioz seized upon the viola's status as an outsider in the world of 19th century music (with which he undoubtedly identified) to fashion a fascinating, highly personalized role for it throughout Harold. Like Berlioz himself, the solo viola finds itself increasingly isolated from the orchestral mainstream. In the first movement, it seems at one with the orchestral depiction of nature – it launches the Harold theme to break out of the rigidity of the opening orchestral fugue, and then in a series of fitful rising figures, struggles to rouse itself and find an appropriate melody for the sonata portion, in which the full orchestra heartily joins, producing a partnership where each stimulates the other in a unified concerted blend of rising excitement and driving momentum. The relationship unravels in the second movement, though – at first an augmented Harold theme blends harmoniously with the pilgrim song, next becomes disruptive with triplet rhythm, and then turns downright annoying, as rapid arpeggiated chords (emulating the guitar Berlioz liked to strum on his mountain walks) are played sul ponticello [near the bridge] for a gratingly nasal, whiny tone that sours the peaceful meditation of the solemn prayer like a rowdy child in church.

The third movement finds the solo viola marginalized, emerging only to play its Harold theme as a distant observer to the intensely human amorous activity being depicted. After the finale leaves reminiscences of the earlier movements behind, the viola is utterly silent, as if Berlioz, having conjured an onslaught of evil and rebellion, finds himself too timid to join in – or, as if, as Hugh Macdonald put it, Berlioz was temperamentally a stranger to his own wishful imaginings. Indeed, it is only heard once more – harmonizing with the off-stage trio's vain attempt to restore a brief breath of serene stability, after which the boiling orchestra leaves them all decisively behind to seal its mutiny.

In his Mémoires, Berlioz blamed the disastrous November 23, 1834 premiere of Harold on the conductor, Girard, who failed to accelerate the end of the first movement, leaving it cold and languid, and then lost count during an encore of the second, calling out for the orchestra to jump to the end. But no matter – the audience seemed far more interested in La Captive, Berlioz's setting of the Hugo poem of the erotic reveries of a harem slave. (The next performance two weeks later gained even lesser notice – the audience was wowed by Chopin, a guest star who played the Romanze from his first piano concerto.) Berlioz was so upset by the premiere that he resolved never to let anyone else conduct any of his work, and to prevent inadequate performances he closely guarded the scores and delayed their publication (Harold appeared only in 1848). Even so, for posterity, Berlioz left some quixotic and obsessively detailed instructions in his score, including placement of the players (the solo viola is to be near the harp), the type of drum-sticks to use on various phrases, the method of rolling tambourines (with the fingers), the number of beats to give in certain measures, and a caution that a gradual crescendo is to extend evenly over 115 measures.

By all accounts, including his own, Berlioz was an exacting conductor who rehearsed each section intensively – from a 1843 concert he cut the finale, finding the trombones incompetent and the violins too few for the sound he required. (The sound Berlioz sought was indeed mammoth – his ideal orchestra would have had 435 players, including 120 violins, 16 horns and 30 harps!) Perhaps in reaction to the emotional egoistic style of Paganini that induced the work, and consistent with his goal of teamwork among all players, Berlioz criticized Wagner's conducting as too free and acclaimed a performance of Harold in which the viola solo had fine rhythmic control and another in which the solo was played by a violinist of the orchestra "who had no pretensions to being a virtuoso." We can only wonder what the impassioned composer would have thought of the diverse recordings we now have of his neglected symphony.

While according to Lance Brunner's compilation, the Symphonie Fantastique boasted four complete recordings (led by Rhené-Bâton, Fried, Weismann and Weingartner) within the first year of the electrical era,

William Primrose

Harold had to wait nearly two more decades for its phonographic debut by the Boston Symphony Orchestra led by Serge Koussevitzky, with William Primrose, violist (now on Biddulph or Naxos CDs). Fortunately, the result was not a mere performance but a true interpretation, bursting with stirring character and probing expression. The first movement is full of tension as the viola leaps ahead, eager for discovery, even as its dark tone suggests lurking menace and the nervous phrasing of the orchestra dispels any sense of a joyous romp in the country. The second movement is unified by a deeply meditative spirituality as the enchanted viola blends effortlessly into the strings' song, and then complements the central section with tender and wistful arpeggios, each shimmering chord and phrase exquisitely molded. The third, too, is cast in a dream-state, with smoothly blended transitions between the disparate sections to modulate the rusticity. The finale brilliantly reflects the composer's historical position by treading a highly appropriate balance between emotional excess and restrained classicism, delving beneath the surface to find a core layer of disturbing terror. The solo viola ardently pleads for a return to stability by stating the earlier themes with pride before reluctantly succumbing to the orchestral entreaty to join the revelry. Climaxes are labored, their gravity defeating any mood of genuine liberation, with distorted bass drum and ear-splitting cymbals (abetted by the stressed recording quality) tearing the sonic fabric with regret and anguish. The trombones stride atop the texture, to reflect the composer's love of the instrument and the importance he attached to its deep but hair-raising sustained tone.

Primrose recalled that he first learned the part at the request of Arturo Toscanini for a 1939 broadcast concert by the NBC Symphony, in which he was the co-principal violist, before embarking on a solo career. That thrilling event is preserved in rather crude sound on a Music & Arts CD. We also have a more sonically flattering, but far less exciting, recording of another 1953 Toscanini/NBC broadcast, with the orchestra's other co-principal violist, Carlton Cooley (BMG). Despite balances that over-emphasize the soloist, who nearly overrides the full orchestra's volume, the 1939 version is far more dynamic, with thrillingly precise ensemble and rhythmic articulation. While the second rendition emphasizes the work's classicism through its unyielding pace (except for acceleration at the very end of the first and last movements), the earlier concert adds an extra measure of energetic abandon and a more subtle flexibility of tempo and phrasing, and its sheer visceral impact is enhanced by a powder-dry acoustic without even a hint of echo or concert-hall atmosphere. The solo playing reflects the diverse aura of each performance, with Primrose wonderfully inflected in the first and Cooley far more reserved in the second.

For quite a while it seemed that Primrose virtually owned Harold on record. He returned as soloist in 1952 with Sir Thomas Beecham conducting his Royal Philharmonic Orchestra (Sony). As with his acclaimed recordings of the Symphonie Fantastique (EMI) and Requiem (BBC), Beecham crafts a patient, graceful and beautifully balanced reading radiating grace and moderation, as if to sublimate his own character to let the music itself gleam (thus, neutrality without the unrelenting rigidity and emotional detachment that drags down the 1953 Toscanini record and that perhaps reached its nadir in the 1977 recording by Lorin Maazel and the Cleveland Orchestra with Robert Vernon on a London LP). While this approach can be lauded as respecting Berlioz, it can also be viewed as a disappointingly aloof refusal to engage the deeply personal passions that fueled composition of this work in the first place. (A 1956 Beecham BBC concert with Frederick Riddle (BBC CD) is slightly brusque and compromised by horrendous solo playing.)

Primrose appeared yet again in 1958 in the first stereo Harold with Charles Munch conducting the Boston Symphony (RCA). Never one to wallow in emotional excess, Munch delivers the fastest Harold on record at a breathless 38 minutes, only in part by omitting the first movement exposition repeat (which would have added about 1¼ minutes). The fleet tempos allow no lingering over details. Thus, the first movement is more a jog than a jaunt through the country and the cruel velocity of the arpeggios in the second really put Primrose to the test (which he passes brilliantly). The third is normally paced, and thus foregoes the shift to a faster gear that usually follows the placid second, while the sweet, vibrato-laden solo discounts the "leather lunged" serenader the composer described. The sharp and clear finale upholds a sense of musical abstraction above the score's characterization of carousing. Balances exaggerate the harp and triangle but the sparkling, transparent recording displays the conductor's characteristic lack of pretension.

Curiously, despite Berlioz's present regard as the quintessential French composer of his time, we have few recordings of Harold by French conductors. Aside from the atypical Munch, one of the very few others hails from the Montreal Symphony led by Charles Dutoit with Pinchas Zuckerman (London) and it's a beauty. At 45 minutes, it's the most leisurely-paced Harold I've heard on record, with exceptionally smooth, blended and patient unfolding of the musical splendors enhanced by Zuckerman's liquid, romanticized phrasing. Yet, to avoid wallowing in thickened textures, the sonority throughout is crisp and textures are luminous. The third movement, in particular, overflows with sheer love for the bucolic scenes that inspired Berlioz.

Another dynasty of Harold recordings to place alongside Primrose's are the three led by Colin Davis. Perhaps recalling the work's origin, Davis's first 1963 recording with the Philharmonia Orchestra (EMI) features the acclaimed Yehudi Menuhin who, like Paganini, was a famed violinist. (Primrose also began his career and made early recordings on the violin, but then abandoned it for the viola.) Perhaps in recognition of Menuhin's status, Davis defers to his soloist's proclivity for inflection. Menuhin's fame appeared to have guided the album's marketing as well, as his name appeared on the LP cover in a far larger font than either conductor or orchestra (and of the performers only Menuhin is identified on the spine labeling). At 44½ minutes, the reading is rather bland and catches fire only at the very end, as urgent accents colorfully fight the otherwise steady underlying rhythm. A Davis-led 1975 Philips LP with the London Symphony Orchestra and Nobuko Imai is sharper, quicker, more rhythmically taut, and less solo-dominated. Davis's most recent Harold is a 2003 London Symphony concert with Tabea Zimmerman (2003, LSO CD) – even faster and more vibrant than in 1975, with each section keenly characterized, but especially remarkable for the extraordinary detail of the recording, from which timbres constantly emerge, fostering a fine appreciation for Berlioz's famed orchestration skills. Although a bit bass-shy, the textural differentiation strongly suggests a sonority that approaches the composer's own aesthetic.

For the full historical treatment, John Eliot Gardiner leads Gerard Caussé and the Orchestre Revolutionnaire et Romantique (1994, Philips CD) in a reading that consciously emulates the performance practices of the composer's time. In notes to their companion CD of the Symphonie Fantastique, Gardiner decried the "pummeling into conformity" by plusher sonorities of modern orchestras that yield a crude substitute for Berlioz's minutely calibrated timbres and textures, and instead justified a blend of authentic and new instruments to achieve an historical reappraisal by reconstructing the characteristic orchestral sound that Berlioz achieved with his supreme instrumentation skill. Their Harold indeed is remarkable, although there are some sonic compromises to modern ears; thus, while the sharp tympani are indeed thrilling, the trombones, which Berlioz regarded as a sonic anchor, emerge as rather light-weight for that role. The sheer sonority adds a vast infusion of telling character, especially evident in the finale, where inner voices hold their own and the shrill piccolo and swirling violins are highlighted rather than blending into the overall mix.

Despite these fine accounts from veteran conductors, let's not forget that Berlioz was only 31 years old when he wrote Harold. More than any other conductor, Leonard Bernstein led the score with a thrilling and highly appropriate youthful exuberance. His 1961 New York Philharmonic recording (Sony) was to have presented yet another solo by Primrose, who apparently fell ill, and so instead features William Lincer, the Philharmonic's principal violist who indeed seems fully integrated into the orchestral fabric and is closely miked for a further sense of immediacy and involvement. Sharp, brash and lean, the result comes close to the feel of period instrumentation, but with a riveting passion absent from most reconstructive attempts. Indeed, the performance is ardently romantic, bristling with energy as Bernstein strains to wrest the utmost fervor out of his forces. Bernstein also left us a 1977 remake with the Orchestre National de France and Donald McInnes (EMI), that is equally exciting, but in a more subtle way. A companion French recording of the Symphonie Fantastique is overtly impetuous and meticulously detailed, but for the first movement of Harold Bernstein takes his cue from Toscanini's careful balance between classical restraint and rebellious invention to portray the duality of Berlioz's conception with constant excited outbursts that arise and subside naturally from within the French orchestral tradition of smooth, mellow sound in which the overall performance is nestled. The result is a fine depiction of the artistic struggle waged within the composer's soul, taking constant flight yet tempered by the aesthetic roots from which he could not fully extract himself. The middle movements provide a serene respite and lead naturally into the finale, which Bernstein depicts as troubled and dark, as if to accentuate the composer's inability to fully identify with his fantasy, but then caps it off with a wild and disconcerting final acceleration, as if to break the mood, shake off any philosophical burdens and arm the youthful Berlioz to prepare for new adventure. The thick sonic ambiance, which not only enhances the "French" sound of the ensemble but contributes so much to its effectiveness, may have arisen in part as an artifact of the quadraphonic format of the original release.

Among other recordings I've enjoyed, the budget-priced version by Rivka Golani and the San Diego Orchestra led by Yaov Talmi (Naxos), while lacking the distinction of those already noted, is surprisingly accomplished and idiomatic. And while at first glance the 37-minute Igor Markevitch recording with Heinz Kirchner seems to trump the others for speed, don't be fooled – the timing is due to a massive cut in the finale (from numbers 46 through 52 in the Breitkopf and Härtel score), despite the composer having written out both identical iterations of the long middle section, as if to insist that the repeat was required. In any event, this is a sturdy, thick, heavily-textured Germanic reading, as would be expected from the Berlin Philharmonic – illustrating Paganini's insistence (seconded by Liszt) that Berlioz was the true artistic heir of Beethoven.

But even amid the excellence of these recordings, my all-time favorite is the overlooked 1953 recording by Hermann Scherchen and the Royal Philharmonic (Westminster LP, briefly available in a 4-CD Tahra box). Sharing much of the sense of sheer adventure of the 1961 Bernstein, and with even greater visceral excitement than the 1939 Toscanini, his approach channels Koussevitzy's huge unbridled personality but adds wild extremes, boasting both the slowest second movement on record and the fastest fourth. After a deceptively placid methodical opening, the mood is shattered with a sudden shift to vast velocity and force, as soloist Frederick Riddle, in far better form than with Beecham, digs into his instrument and tosses off his figures with striking staccato amid orchestral snarls. The middle movements boast distinctive, humanizing touches – the second evolves into a profoundly religious rapture, as gingerly phrasing gradually gains confidence with the lengthening lines, and the third evokes the amateurism of Berlioz's serenader through labored pacing and choppy accompaniment. But it's in the finale that Scherchen pulls out all the stops, violently contrasting the sweet moments of repose with wild breakneck climactic tempos, fully conveying the volatility of the composer's psyche, torn between the stability of social expectations and his fevered imagination. Scherchen saves his biggest surprise for the very end, though – having led us to expect a truly ferocious finish, he catches us off guard with a thoroughly restrained final climax, as if to remind us that, after all, this is concert-hall music nestled in the conventions of a bygone era. Of all the performances on record, Scherchen's comes closest to conjuring for modern listeners the sheer shock and wonderment Berlioz's audiences must have felt when first encountering his music.

The many sources for information about Harold In Italy and Berlioz in general begin with the composer's own Mémoires, available in several editions and translations. Among biographies, I found those by D. Kern Holomon (Harvard University Press, 1989), David Cairns (University of California, 1999) and Hugh Macdonald (J. M. Dent & Sons, 1982) especially informative. Macdonald also authored a BBC Music Guide to Berlioz's Orchestral Music (University of Washington Press 1969) and Cairns contributed a valuable article on Berlioz to The Symphony compendium edited by Robert Simpson (Penguin, 1966). The Cambridge Companion to Berlioz (edited by Peter Bloom, Cambridge University Press 2000) includes articles by Julian Preston on genre, Pierre Citron on the Mémoires and Joel-Marie Fouquet on the Grande Traité d'instrumentation et d'orchestration. Jacques Barzin's Berlioz and the Romantic Century (Little, Brown & Company 1950) is an extraordinary and extensive analysis of the composer in the context of his psyche and times. Paul Henry Lang's sweeping Music in Western Civilization (W. W. Norton & Company 1941) is deeply insightful and personal. Above all else, the full orchestral score (reprinted by Dover in the same volume as the Symphonie Fantastique, 1984) displays the extraordinary complexity of Berlioz's conception. This article is indebted to all of these, along with liner notes to the LPs and CDs I've cited.

Copyright 2008 by Peter Gutmann