Context:
fun
 of
 that
 story
 was
 
going
 to
 make
 that
 story
 live
 
longer
.
 It
'
s
 so
 easy
 to
 accident
ally
 engage
 in
 
problematic
 behavior
 or
 behavior
 that
 
could
 harm
 direct
ly
 or
 in
direct
ly
 —
 we
 don
'
t
 ever
 have
 to
 see
 what
 the
 full
 re
percussion
s
 of
 our
 behavior
s
 are
 (
online
).
 I
 think
 
thinking
 about
 
those
 more
 
everyday
 behavior
s
 as
 important
,
 just
 as
 it
'
s
 important
 to
 think
 about
 the
 true
 bad
 
actors
 —
 the
 
ones
 who
 
really
 want
 to
 
cause
 harm
.
 We
 don
'
t
 have
 control
 of
 what
 the
 bad
 
guys
 do
,
 but
 we
 do
 have
 control
 over
 our
 own
 behavior
s
.
 2.
 This
 is
 
a
 b
igger
 problem
 than
 people
 
having
 the
 wrong
 tone
 in
 how
 the
y
 speak
 to
 one
 
another
.
 WP
:
 I
 think
 that
 civil
ity
 is
 sort
 of
 
a
 trick
y
 concept
.
 Of
ten
times
 civil
ity
 get
s
 used
 as
 
a
 way
 to
 sort
 of
 silent
 dissen
ting
 
speech
.
 That
 is
,
 
someone
 rais
es
 
their
 voice
 and
 you
 say
,
 "
You
'
re
 
being
 un
civil
."
 "
I
 don
'
t
 want
 to
 listen
 to
 your
 grie
vance
.
 So
 be
 civil
."
 So
 it
'
s
 not
 civil
ity
 that
 is
 what
'
s
 la
cking
 online
.
 I
 think
 it
'
s
 
ethic
s
.
 I
 think
 it
'
s
 
thinking
 about
 the
 full
 
i
mplications
 of
 our
 
everyday
 behavior
.
 3.
 Post
ing
 about
 something
 just
 to
 make
 fun
 of
 it
?
 You
'
re
 part
 of
 the
 problem
.
 WP
:
 A
 lot
 of
 people
 on
 the
 left
 —
 
a
 lot
 of
 
academic
s
,
 even
 —
 
spent
 
their
 summer
 posting
 an
 
apparent
ly
 sincer
e
 
QA
non
 conspira
cy
 
theory
 in
 order
 to
 make
 fun
 of
 it
.
 You
 know
,
 
simply
 by
 engag
ing
 with
 it
 to
 make
 fun
 of
 it
 or
 to
 c
ritique
 it
 or
 any
 of
 that
,
 it
 
makes
 the
 story
 b
igger
,
 and
 wide
n
s
 the
 
gulf
 
between
 people
 who
 
could
 potential
ly
 have
 meaning
ful
 conversation
s
.
 On
 the
 
fring
e
,
 people
 are
 
going
 to
 do
 what
 the
y
'
re
 
going
 to
 do
.
 But
 
a
 lot
 of
 people
 —
 moderat
e
 
conservative
s
 —
 
would
 be
 so
 ange
red
 by
 the
 liberal
 impuls
e
 to
 call
 Trump
 supporter
s
 stupid
 and
 crazy
 and
 all
 
those
 de
rog
atory
 words
 that
 (
they
 are
)
 less
 like
ly
 to
 have
 moment
s
 
where
 you
 can
 potential
ly
 find
 common
 ground
.
 4.
 Get
ting
 put
 into
 
us
 vers
us
 them
 si
los
 dis
courage
s
 thought
ful
 dis
course
.
 WP
:
 I
'
m
 not
 say
ing
 that
 you
 need
Input:
ly
 
when
 forward
ed
 in
 response
 to
 
critical
,
 dissen
ting
 
speech
.
 Side
ste
pping
 the
 
content
 of
 
a
 c
ritique
 in
 order
 to
 police
 the
 tone
 of
 that
 c
ritique
—
a
 
strategy
 employ
ed
 with
 particular
 
vigor
 
during
 the
 K
avana
ugh
 hear
ings
,
 and
 
which
 frequent
ly
 factor
s
 into
 hand
-
w
ring
ing
 over
 anti
-
racist
 activ
ism
—
serves
 to
 
false
y
 equ
ate
 civil
ity
 with
 polite
ness
,
 and
 polite
ness
 with
 the
 
democratic
 ideal
.
 In
 short
:
 you
 are
 
being
 civil
 
when
 you
 don
’
t
 
ruffle
 my
 
feather
s
,
 
which
 is
 to
 say
,
 
when
 I
 don
’
t
 have
 to
 hear
 your
 grie
vance
.
 Be
sides
 
their
 tendenc
y
 to
 be
 adopt
ed
 as
 bad
 
faith
,
 rhe
tor
ical
 
s
leigh
t
s
-
of
-
hand
,
 
calls
 for
 civil
ity
 have
 
another
,
 per
haps
 more
 in
sidi
ous
,
 con
sequence
:
 de
f
lect
ing
 b
lame
.
 It
’
s
 
e
verybody
 else
’
s
 behavior
,
 the
y
’
re
 the
 
ones
 who
 need
 to
 start
 
acting
 right
.
 
They
’
re
 the
 
ones
 who
 need
 to
 control
 
themselves
.
 In
 
these
 
instance
s
,
 “
We
 need
 to
 
restore
 civil
ity
”
 
become
s
 an
 
exercise
 in
 finger
 point
ing
.
 You
’
re
 the
 one
 who
 is
n
’
t
 
being
 civil
.
 Inde
ed
,
 the
 above
 
NPR
 survey
 
explicit
ly
 
asked
 
respondent
s
 to
 identif
y
 who
 was
 to
 b
lame
 for
 the
 
lack
 of
 civil
ity
 in
 Washington
,
 with
 four
 possible
 choice
s
:
 President
 Trump
,
 
Republican
s
 in
 Congress
,
 Democrat
s
 in
 Congress
,
 or
 the
 media
.
 Who
se
 
fault
 is
 it
:
 this
 is
 how
 the
 civil
ity
 question
 tend
s
 to
 be
 fram
ed
.
 Ethi
c
s
 do
 not
 mean
 
keeping
 your
 voice
 down
.
 Ethi
c
s
 do
 not
 mean
 
keeping
 
feather
s
 un
ruffle
d
.
 Ethi
c
s
 mean
 
taking
 full
 and
 un
qualifie
d
 
responsibility
 for
 the
 things
 you
 
choose
 to
 do
 and
 say
.
 We
 
certain
ly
 
maintain
 that
 the
 behavior
 of
 others
 can
 be
 
a
 problem
,
 or
 out
right
 
dangerous
.
 We
 
certain
ly
 
maintain
 that
 some
 people
 need
 to
 control
 
themselves
,
 particular
ly
 
given
 the
 
increas
Targets:
ingly Idx: 29747, Δloglikelihood: -0.139
  Idx: 259, Δloglikelihood: 0.029
glar Idx: 106359, Δloglikelihood: 1.980
ing Idx: 347, Δloglikelihood: 0.032
 link Idx: 5275, Δloglikelihood: -1.653
  Idx: 259, Δloglikelihood: -0.019
between Idx: 4964, Δloglikelihood: -0.001
 violent Idx: 57219, Δloglikelihood: -1.072
  Idx: 259, Δloglikelihood: 0.611
political Idx: 28735, Δloglikelihood: -0.936
 rhe Idx: 87309, Δloglikelihood: -0.590
tori Idx: 9727, Δloglikelihood: 0.008
c Idx: 297, Δloglikelihood: -0.000
 and Idx: 305, Δloglikelihood: 0.025
 violent Idx: 57219, Δloglikelihood: 1.918
 action Idx: 13072, Δloglikelihood: 1.270
. Idx: 260, Δloglikelihood: 0.181
 T Idx: 366, Δloglikelihood: 0.421
hose Idx: 51571, Δloglikelihood: -0.046
 who Idx: 1866, Δloglikelihood: -0.452
 trade Idx: 23853, Δloglikelihood: 0.780
 in Idx: 281, Δloglikelihood: 0.031
 an Idx: 461, Δloglikelihood: -0.453
tagon Idx: 147779, Δloglikelihood: -2.626
ism Idx: 6410, Δloglikelihood: 0.084
, Idx: 261, Δloglikelihood: 0.169
 in Idx: 281, Δloglikelihood: -0.051
  Idx: 259, Δloglikelihood: -0.563
manipulation Idx: 157440, Δloglikelihood: 3.039
, Idx: 261, Δloglikelihood: 0.203
 in Idx: 281, Δloglikelihood: 0.696
 symboli Idx: 111137, Δloglikelihood: 0.965
c Idx: 297, Δloglikelihood: -0.037
  Idx: 259, Δloglikelihood: 0.176
violence Idx: 60641, Δloglikelihood: -0.098
 and Idx: 305, Δloglikelihood: 0.713
  Idx: 259, Δloglikelihood: -0.097
physical Idx: 28223, Δloglikelihood: 0.183
  Idx: 259, Δloglikelihood: -0.505
violence Idx: 60641, Δloglikelihood: -0.186
, Idx: 261, Δloglikelihood: 0.009
  Idx: 259, Δloglikelihood: 0.016
warrant Idx: 136377, Δloglikelihood: -0.336
 special Idx: 5860, Δloglikelihood: 1.246
, Idx: 261, Δloglikelihood: -0.704
 un Idx: 335, Δloglikelihood: -0.768
fl Idx: 7410, Δloglikelihood: -0.478
inch Idx: 29168, Δloglikelihood: 0.003
ing Idx: 347, Δloglikelihood: 0.007
 con Idx: 450, Δloglikelihood: 0.849
dem Idx: 2930, Δloglikelihood: -0.006
nation Idx: 72920, Δloglikelihood: 0.004
. Idx: 260, Δloglikelihood: -0.046
 But Idx: 4837, Δloglikelihood: 0.241
  Idx: 259, Δloglikelihood: 0.129
few Idx: 6924, Δloglikelihood: -0.750
 of Idx: 304, Δloglikelihood: -0.242
  Idx: 259, Δloglikelihood: 0.086
us Idx: 438, Δloglikelihood: 0.176
 are Idx: 418, Δloglikelihood: 0.099
 tru Idx: 10789, Δloglikelihood: 1.089
ly Idx: 484, Δloglikelihood: 0.001
 b Idx: 330, Δloglikelihood: 0.366
lame Idx: 69621, Δloglikelihood: 0.020