THE MECHANICAL
FALLACY 109
whethertheidealisas rationalandconsistent
asit
sounds.
In the first place,it is clear that
the vivid con-
structive properties ofabuilding,in sofarasthey
areeffectivelyconstructive,must
existas/ac^5i,- The
securityof
the
building,and
hencealsoofanyartistic
valueitmaypossess,dependsonthis
;
andasupport
whichseemedto
be
adequatetoitsload,butactually
wasnot,would, asconstruction,bewrong. Butin
sofar
as
they
arej\a^d,theymustexist
asappear-
ances.. It is theeffect which the constructive pro-
pertiesmakeontheeye,andnotthescientificfacts
thatmaybe
intellectuallydiscoverable
aboutthem,
which alone can determine their vividness. Con-
struction, it may be granted, is always, or nearly
always, insomesense, ourconcern,butnot always
in the same sense. The two requirements
which
architecture so far evidently has are constructive
integrity in fact, and constructive vividness in
appearance.
Now, what our scientific critics
have
takenforgranted,
isthatbecausethese
tworequire-
ments have
sometimes beeil satisfied at the same
moment, andby
the same means, no
otherwayof
satisfying
themis permissible. Buttherehas been
nonecessityshown
thusfar,norisit
easytoimagine
one,for
insistingthatthesetwoqualificationsshould
alwaysbeinterdependent,
andthatboth
mustinvari-
ablybesatisfied
atasinglestroke. Theirvaluein