<s>
Planner	B-Application
(	O
often	O
seen	O
in	O
publications	O
as	O
"	O
PLANNER	B-Application
"	O
although	O
it	O
is	O
not	O
an	O
acronym	O
)	O
is	O
a	O
programming	O
language	O
designed	O
by	O
Carl	O
Hewitt	O
at	O
MIT	O
,	O
and	O
first	O
published	O
in	O
1969	O
.	O
</s>
<s>
First	O
,	O
subsets	O
such	O
as	O
Micro-Planner	O
and	O
Pico-Planner	O
were	O
implemented	O
,	O
and	O
then	O
essentially	O
the	O
whole	O
language	O
was	O
implemented	O
as	O
Popler	O
by	O
Julian	O
Davies	O
at	O
the	O
University	O
of	O
Edinburgh	O
in	O
the	O
POP-2	B-Language
programming	O
language	O
.	O
</s>
<s>
Derivations	O
such	O
as	O
QA4	O
,	O
Conniver	O
,	O
QLISP	O
and	O
Ether	O
(	O
see	O
scientific	B-Application
community	I-Application
metaphor	I-Application
)	O
were	O
important	O
tools	O
in	O
artificial	B-Application
intelligence	I-Application
research	I-Application
in	O
the	O
1970s	O
,	O
which	O
influenced	O
commercial	O
developments	O
such	O
as	O
Knowledge	O
Engineering	O
Environment	O
(	O
KEE	O
)	O
and	O
Automated	O
Reasoning	O
Tool	O
(	O
ART	O
)	O
.	O
</s>
<s>
The	O
two	O
major	O
paradigms	O
for	O
constructing	O
semantic	O
software	O
systems	O
were	O
procedural	B-Application
and	O
logical	B-Language
.	O
</s>
<s>
Lisp	B-Language
[	O
McCarthy	O
et	O
al	O
.	O
</s>
<s>
The	O
logical	B-Language
paradigm	O
was	O
epitomized	O
by	O
uniform	O
proof	O
procedure	O
resolution-based	O
derivation	O
(	O
proof	O
)	O
finders	O
[	O
Robinson	O
1965 ]	O
.	O
</s>
<s>
According	O
to	O
the	O
logical	B-Language
paradigm	O
it	O
was	O
“	O
cheating	O
”	O
to	O
incorporate	O
procedural	B-Application
knowledge	O
[	O
Green	O
1969 ]	O
.	O
</s>
<s>
Converted	O
everything	O
to	O
clausal	B-Application
form	I-Application
.	O
</s>
<s>
Converting	O
all	O
information	O
to	O
clausal	B-Application
form	I-Application
is	O
problematic	O
because	O
it	O
hides	O
the	O
underlying	O
structure	O
of	O
the	O
information	O
.	O
</s>
<s>
Then	O
used	O
resolution	O
to	O
attempt	O
to	O
obtain	O
a	O
proof	O
by	O
contradiction	O
by	O
adding	O
the	O
clausal	B-Application
form	I-Application
of	O
the	O
negation	O
of	O
the	O
theorem	O
to	O
be	O
proved	O
.	O
</s>
<s>
Planner	B-Application
was	O
a	O
kind	O
of	O
hybrid	O
between	O
the	O
procedural	B-Application
and	O
logical	B-Language
paradigms	O
because	O
it	O
combined	O
programmability	O
with	O
logical	B-Language
reasoning	O
.	O
</s>
<s>
Planner	B-Application
featured	O
a	O
procedural	B-Application
interpretation	O
of	O
logical	B-Language
sentences	O
where	O
an	O
implication	O
of	O
the	O
form	O
can	O
be	O
procedurally	O
interpreted	O
in	O
the	O
following	O
ways	O
using	O
pattern-directed	O
invocation	O
:	O
</s>
<s>
In	O
this	O
respect	O
,	O
the	O
development	O
of	O
Planner	B-Application
was	O
influenced	O
by	O
natural	O
deductive	O
logical	B-Language
systems	O
(	O
especially	O
the	O
one	O
by	O
Frederic	B-Application
Fitch	I-Application
 [ 1952 ] 	O
)	O
.	O
</s>
<s>
A	O
subset	O
called	O
Micro-Planner	O
was	O
implemented	O
by	O
Gerry	O
Sussman	O
,	O
Eugene	O
Charniak	O
and	O
Terry	O
Winograd	O
[	O
Sussman	O
,	O
Charniak	O
,	O
and	O
Winograd	O
1971 ]	O
and	O
was	O
used	O
in	O
Winograd	O
's	O
natural-language	O
understanding	O
program	O
SHRDLU	O
,	O
Eugene	O
Charniak	O
's	O
story	O
understanding	O
work	O
,	O
Thorne	O
McCarty	O
's	O
work	O
on	O
legal	O
reasoning	O
,	O
and	O
some	O
other	O
projects	O
.	O
</s>
<s>
This	O
generated	O
a	O
great	O
deal	O
of	O
excitement	O
in	O
the	O
field	O
of	O
AI	B-Application
.	O
</s>
<s>
It	O
also	O
generated	O
controversy	O
because	O
it	O
proposed	O
an	O
alternative	O
to	O
the	O
logic	B-Language
approach	O
that	O
had	O
been	O
one	O
of	O
the	O
mainstay	O
paradigms	O
for	O
AI	B-Application
.	O
</s>
<s>
At	O
SRI	O
International	O
,	O
Jeff	O
Rulifson	O
,	O
Jan	O
Derksen	O
,	O
and	O
Richard	O
Waldinger	O
developed	O
QA4	O
which	O
built	O
on	O
the	O
constructs	O
in	O
Planner	B-Application
and	O
introduced	O
a	O
context	O
mechanism	O
to	O
provide	O
modularity	O
for	O
expressions	O
in	O
the	O
database	O
.	O
</s>
<s>
Earl	O
Sacerdoti	O
and	O
Rene	O
Reboh	O
developed	O
QLISP	O
,	O
an	O
extension	O
of	O
QA4	O
embedded	O
in	O
INTERLISP	B-Language
,	O
providing	O
Planner-like	O
reasoning	O
embedded	O
in	O
a	O
procedural	B-Application
language	I-Application
and	O
developed	O
in	O
its	O
rich	O
programming	O
environment	O
.	O
</s>
<s>
So	O
Planner	B-Application
adopted	O
some	O
efficiency	O
expedients	O
including	O
the	O
following	O
:	O
</s>
<s>
A	O
closed-world	B-Application
assumption	I-Application
could	O
be	O
implemented	O
by	O
conditionally	O
testing	O
whether	O
an	O
attempt	O
to	O
prove	O
a	O
goal	O
exhaustively	O
failed	O
.	O
</s>
<s>
Later	O
this	O
capability	O
was	O
given	O
the	O
misleading	O
name	O
"	O
negation	B-Application
as	I-Application
failure	I-Application
"	O
because	O
for	O
a	O
goal	O
it	O
was	O
possible	O
to	O
say	O
:	O
"	O
if	O
attempting	O
to	O
achieve	O
exhaustively	O
fails	O
then	O
assert	O
.	O
"	O
</s>
<s>
Gerry	O
Sussman	O
,	O
Eugene	O
Charniak	O
,	O
Seymour	O
Papert	O
and	O
Terry	O
Winograd	O
visited	O
the	O
University	O
of	O
Edinburgh	O
in	O
1971	O
,	O
spreading	O
the	O
news	O
about	O
Micro-Planner	O
and	O
SHRDLU	O
and	O
casting	O
doubt	O
on	O
the	O
resolution	O
uniform	O
proof	O
procedure	O
approach	O
that	O
had	O
been	O
the	O
mainstay	O
of	O
the	O
Edinburgh	O
Logicists	O
.	O
</s>
<s>
At	O
the	O
University	O
of	O
Edinburgh	O
,	O
Bruce	O
Anderson	O
implemented	O
a	O
subset	O
of	O
Micro-Planner	O
called	O
PICO-PLANNER	O
(	O
Anderson	O
1972	O
)	O
and	O
Julian	O
Davies	O
(	O
1973	O
)	O
implemented	O
essentially	O
all	O
of	O
Planner	B-Application
.	O
</s>
<s>
According	O
to	O
Donald	O
MacKenzie	O
,	O
Pat	O
Hayes	O
recalled	O
the	O
impact	O
of	O
a	O
visit	O
from	O
Papert	O
to	O
Edinburgh	O
,	O
which	O
had	O
become	O
the	O
"	O
heart	O
of	O
artificial	B-Application
intelligence	I-Application
's	O
Logicland	O
,	O
"	O
according	O
to	O
Papert	O
's	O
MIT	O
colleague	O
,	O
Carl	O
Hewitt	O
.	O
</s>
<s>
These	O
tensions	O
were	O
exacerbated	O
when	O
the	O
UK	O
Science	O
Research	O
Council	O
commissioned	O
Sir	O
James	O
Lighthill	O
to	O
write	O
a	O
report	O
on	O
the	O
AI	B-Application
research	I-Application
situation	O
in	O
the	O
UK	O
.	O
</s>
<s>
Pat	O
Hayes	O
visited	O
Stanford	O
where	O
he	O
learned	O
about	O
Planner	B-Application
.	O
</s>
<s>
When	O
he	O
returned	O
to	O
Edinburgh	O
,	O
he	O
tried	O
to	O
influence	O
his	O
friend	O
Bob	O
Kowalski	O
to	O
take	O
Planner	B-Application
into	O
account	O
in	O
their	O
joint	O
work	O
on	O
automated	O
theorem	O
proving	O
.	O
</s>
<s>
He	O
carefully	O
studied	O
Planner.	O
”	O
according	O
to	O
Bruynooghe	O
,	O
Pereira	O
,	O
Siekmann	O
,	O
and	O
van	O
Emden	O
 [ 2004 ] 	O
.	O
</s>
<s>
Kowalski	O
 [ 1988 ] 	O
states	O
"	O
I	O
can	O
recall	O
trying	O
to	O
convince	O
Hewitt	O
that	O
Planner	B-Application
was	O
similar	O
to	O
SL-resolution.	O
"	O
</s>
<s>
But	O
Planner	B-Application
was	O
invented	O
for	O
the	O
purposes	O
of	O
the	O
procedural	B-Application
embedding	O
of	O
knowledge	O
and	O
was	O
a	O
rejection	O
of	O
the	O
resolution	O
uniform	O
proof	O
procedure	O
paradigm	O
.	O
</s>
<s>
Colmerauer	O
and	O
Roussel	O
recalled	O
their	O
reaction	O
to	O
learning	O
about	O
Planner	B-Application
in	O
the	O
following	O
way	O
:	O
</s>
<s>
It	O
was	O
at	O
this	O
time	O
that	O
we	O
learned	O
of	O
the	O
existence	O
of	O
Carl	O
Hewitt	O
’s	O
programming	O
language	O
,	O
Planner	B-Application
[	O
Hewitt	O
,	O
1969 ]	O
.	O
</s>
<s>
The	O
lack	O
of	O
formalization	O
of	O
this	O
language	O
,	O
our	O
ignorance	O
of	O
Lisp	B-Language
and	O
,	O
above	O
all	O
,	O
the	O
fact	O
that	O
we	O
were	O
absolutely	O
devoted	O
to	O
logic	B-Language
meant	O
that	O
this	O
work	O
had	O
little	O
influence	O
on	O
our	O
later	O
research.	O
"	O
</s>
<s>
In	O
the	O
fall	O
of	O
1972	O
,	O
Philippe	O
Roussel	O
implemented	O
a	O
language	O
called	O
Prolog	B-Language
(	O
an	O
abbreviation	O
for	O
PROgrammation	O
en	O
LOGique	O
–	O
French	O
for	O
"	O
programming	O
in	O
logic	B-Language
"	O
)	O
.	O
</s>
<s>
Prolog	B-Language
programs	O
are	O
generically	O
of	O
the	O
following	O
form	O
(	O
which	O
is	O
a	O
special	O
case	O
of	O
the	O
backward-chaining	O
in	O
Planner	B-Application
)	O
:	O
</s>
<s>
Prolog	B-Language
duplicated	O
the	O
following	O
aspects	O
of	O
Micro-Planner	O
:	O
</s>
<s>
Giving	O
up	O
on	O
the	O
completeness	O
paradigm	O
that	O
had	O
characterized	O
previous	O
work	O
on	O
theorem	O
proving	O
and	O
replacing	O
it	O
with	O
the	O
programming	O
language	O
procedural	B-Application
embedding	O
of	O
knowledge	O
paradigm	O
.	O
</s>
<s>
Prolog	B-Language
also	O
duplicated	O
the	O
following	O
capabilities	O
of	O
Micro-Planner	O
which	O
were	O
pragmatically	O
useful	O
for	O
the	O
computers	O
of	O
the	O
era	O
because	O
they	O
saved	O
space	O
and	O
time	O
:	O
</s>
<s>
The	O
way	O
that	O
Planner	B-Application
established	O
that	O
something	O
was	O
provable	O
was	O
to	O
successfully	O
attempt	O
it	O
as	O
a	O
goal	O
and	O
the	O
way	O
that	O
it	O
establish	O
that	O
something	O
was	O
unprovable	O
was	O
to	O
attempt	O
it	O
as	O
a	O
goal	O
and	O
explicitly	O
fail	O
.	O
</s>
<s>
Planner	B-Application
also	O
had	O
a	O
construct	O
which	O
succeeded	O
if	O
failed	O
,	O
which	O
gave	O
rise	O
to	O
the	O
“	O
Negation	B-Application
as	I-Application
Failure	I-Application
”	O
terminology	O
in	O
Planner	B-Application
.	O
</s>
<s>
Use	O
of	O
the	O
Unique	O
Name	O
Assumption	O
and	O
Negation	B-Application
as	I-Application
Failure	I-Application
became	O
more	O
questionable	O
when	O
attention	O
turned	O
to	O
Open	O
Systems	O
[	O
Hewitt	O
and	O
de	O
Jong	O
1983	O
,	O
Hewitt	O
1985	O
,	O
Hewitt	O
and	O
Inman	O
1991 ]	O
.	O
</s>
<s>
The	O
following	O
capabilities	O
of	O
Micro-Planner	O
were	O
omitted	O
from	O
Prolog	B-Language
:	O
</s>
<s>
Logical	B-Language
negation	O
,	O
e.g.	O
,	O
.	O
</s>
<s>
Prolog	B-Language
did	O
not	O
include	O
negation	O
in	O
part	O
because	O
it	O
raises	O
implementation	O
issues	O
.	O
</s>
<s>
Consider	O
for	O
example	O
if	O
negation	O
were	O
included	O
in	O
the	O
following	O
Prolog	B-Language
program	O
:	O
</s>
<s>
The	O
above	O
program	O
would	O
be	O
unable	O
to	O
prove	O
even	O
though	O
it	O
follows	O
by	O
the	O
rules	O
of	O
mathematical	O
logic	B-Language
.	O
</s>
<s>
This	O
is	O
an	O
illustration	O
of	O
the	O
fact	O
that	O
Prolog	B-Language
(	O
like	O
Planner	B-Application
)	O
is	O
intended	O
to	O
be	O
a	O
programming	O
language	O
and	O
so	O
does	O
not	O
(	O
by	O
itself	O
)	O
prove	O
many	O
of	O
the	O
logical	B-Language
consequences	O
that	O
follow	O
from	O
a	O
declarative	O
reading	O
of	O
its	O
programs	O
.	O
</s>
<s>
The	O
work	O
on	O
Prolog	B-Language
was	O
valuable	O
in	O
that	O
it	O
was	O
much	O
simpler	O
than	O
Planner	B-Application
.	O
</s>
<s>
However	O
,	O
as	O
the	O
need	O
arose	O
for	O
greater	O
expressive	O
power	O
in	O
the	O
language	O
,	O
Prolog	B-Language
began	O
to	O
include	O
many	O
of	O
the	O
capabilities	O
of	O
Planner	B-Application
that	O
were	O
left	O
out	O
of	O
the	O
original	O
version	O
of	O
Prolog	B-Language
.	O
</s>
