James Webster, in his article entitled "On the absence of Keyboard Continuo in Haydn's Symphonies" makes the case that
Haydn did not intend a good portion of his middle and late period symphonies to be peformed with keyboard
continuo. Webster mentions that a number of Haydn's early symphonies needed a keyboard continuo to "fill out the
harmonies," and to "keep the
ensemble together," especially during very slow movements. The case for keyboard continuo in the early symphonies is also made by the fact that very often Haydn would conduct the symphonies from the
cembalo itself.
According to Webster, the middle and late symphonies show no sign of keyboard continuo being present, or indeed even optional. He argues that there is no rule demanding the use of keyboard continuo in the symphony, inside or outside of Austria at the time; some did, some didn't. Webster notes that those that did often left some trace of evidence behind, either in the printed work itself or in the documentation referring to it (e.g., correspondence, reviews, etc). He states that there is no evidence for keyboard continuo in these later Haydn symphonies, even referential evidence in the form of "anecdotal testimony," which he believes strengthens his case for the absence of such a practice. Webster does ask for an indulgence in his argument and states that it is impossible to prove a negative. With so much of Haydn's music surviving to the present day, he argues, it would seem almost impossible that only the keyboard continuo parts are missing. He also brings up the interesting point that often Haydn may have conducted the orchestra at Esterhazy from the violin, on which Haydn was most proficient, thereby negating any need to keep the orchestra together by means of a keyboard instrument. Webster does defer to Haydn's keyboard virtuosity, stating that he could have played continuo even without the figures written down, and that no keyboard player other than Haydn himself was ever employed at Esterhazy. This argument is weakened by what Webster views as the ambiguity of the phrase "first violin," which at the time could have meant an orchestra leader or just a member of the string section. An interesting piece of evidence, although by no means conclusive, is the written marking of "Basso Continuo" by Haydn in his score to "Le Midi," however this may be a mistake, since Haydn often wrote just "Basso" for parts that we would today call "Basso Continuo." This may indeed hold a different meaning to modern musicians than it did for Haydn and his contemporaries. Also, since scholarship gives us a good idea of exactly who played at this performance, and Haydn himself was listed as playing the violin, the past assertion that no other keyboard player was hired gives one the freedom to assume that no keyboard continuo was used.
Webster's final argument is that of style. He again mentions "Le Midi" and the "Farewell" Symphony each with solo sections for virtually every instrument group, but there is no obbligato solo for the keyboard in either piece. Webster closes with a reassertion of Haydn's early work as "immature" and therefore in need of "filling out" by the keyboard.