that he was able to return forthwith to Denmark. Now this operation of
accident is a fact, and a prominent fact, of human life. To exclude it
_wholly_ from tragedy, therefore, would be, we may say, to fail in
truth. And, besides, it is not merely a fact. That men may start a
course of events but can neither calculate nor control it, is a _tragic_
fact. The dramatist may use accident so as to make us feel this; and
there are also other dramatic uses to which it may be put. Shakespeare
accordingly admits it. On the other hand, any _large_ admission of
chance into the tragic sequence[4] would certainly weaken, and might
destroy, the sense of the causal connection of character, deed, and
catastrophe. And Shakespeare really uses it very sparingly. We seldom
find ourselves exclaiming, 'What an unlucky accident!' I believe most
readers would have to search painfully for instances. It is, further,
frequently easy to see the dramatic intention of an accident; and some
things which look like accidents have really a connection with
character, and are therefore not in the full sense accidents. Finally, I
believe it will be found that almost all the prominent accidents occur
when the action is well advanced and the impression of the causal
sequence is too firmly fixed to be impaired.

Thus it appears that these three elements in the 'action' are
subordinate, while the dominant factor consists in deeds which issue
from character. So that, by way of summary, we may now alter our first
statement, 'A tragedy is a story of exceptional calamity leading to the
death of a man in high estate,' and we may say instead (what in its turn
is one-sided, though less so), that the story is one of human actions
producing exceptional calamity and ending in the death of such a man.[5]

* * * * *