Attachment to land: The case of the land of Israel for American and Israeli Jews and the role of contagion

Judgment and Decision Making, Vol. 3, No. 4, April 2008, pp. 325–334
Attachment to land: The case of the land of Israel for American and
Israeli Jews and the role of contagion
Paul Rozin? and Sharon Wolf
Department of Psychology
University of Pennsylvania
Abstract
This is a ?rst study on attachment to national and sacred land and land as a protected value. A measure of attachment
to the land of Israel is developed and administered to two groups, Jewish college students in Israel and the United States.
Levels of land attachment are high and not signi?cantly different in the two groups, with a great deal of variation. Land
may become more important through being inhabited by a group over centuries. This is a positive contagion effect,
and is opposed in some cases by negative contagion produced when the “enemies” live on the land for some period of
time. We demonstrate a signi?cant correlation of positive contagion sensitivity with attachment to the land of Israel.
Unlike many other cases of the interaction of positive and negative contagion, negative contagion does not overwhelm
positive contagion in the domain of land attachment. We also present evidence for linkages between political positions,
religiosity, importance of Israel, Arab aversion, and vulnerability of Israel with attachment to land, but these do not fully
account for the contagion effects. A number of signi?cant differences between Israelis and Americans are described.
Keywords: political, ethnic, land, attachment, tradability, contagion.
1 Introduction
attachment “system” may have been extended to sacred
land, and later to national land, by a process of cultural
People often care deeply about their personal land, for ex-
preadaptation (Mayr, 1960; Rozin, 1999).
ample, the land which their family has owned and lived
Land is often thought of as untradable, which would
on for decades to centuries. This may be linked through
not be the case if it was just a source of resources. The
evolution to territorial defense. But people also seem
French word terroir captures a broader perspective, and
to care about land that is relevant to their group, rather
refers to the land including its human capital and cultural
than their own person or family. This land “attachment”
history. None of these strong attachments has been sys-
may be considered in two distinct ways: an attachment
tematically studied by psychologists.
to “place,” that is, speci?c locations of particular histor-
ical signi?cance, and attachment to “space”, a more ab-
Land plays a central role in many current political and
stract type of attachment to a larger parcel of land that
ethnopolitical con?icts, including Northern Ireland, Iraq,
has some sociopolitical signi?cance, such as the land de-
Taiwan, Palestine/Israel and Sri Lanka. States very rarely
?ned by the borders of a country (Deudney, 1996). The
give up land voluntarily, even if that land is inhabited by
“place” attachment (“sacred land”) probably has a history
people who are despised by, and despise, the host country.
of thousands of years, whereas the attachment to land as
This fact is emphasized by Lustick (1993), as he explores
“space” is much more recent. National land is the newest
three of the very few cases in which a country “voluntar-
aspect of land, since nations and states are relatively new
ily” gave up land in the 20th century.
inventions in human history (Weber, 1977). As nations
The concept of protected values (Baron & Spranca,
and states developed, borders came to de?ne territories,
1997) is relevant to land attachment. Protected values
and were defended and expanded. The personal land
are those “unfungible” values or entities which people
resist trading for anything. It is morally offensive for
?We thank Sonia Roccas and Ilana Ritov for cooperation in ob-
taining the Israeli sample, Ian Lustick for discussions of relevant is-
at least some people to even consider exchanging one’s
sues about Israel-Palestine in forming our measures, and Jonathan
children, one’s religion, and one’s land; such potential
Baron, Ian Lustick, Ifat Maoz, Clark McCauley, Ilana Ritov and So-
exchanges have also been described as “taboo tradeoffs”
nia Roccas, for constructive comments on the manuscript. Send cor-
involving sacred values (Fiske & Tetlock, 1997; Tetlock
respondence to: Paul Rozin, Department of Psychology, University of
Pennsylvania, 3720 Walnut St., Philadelphia, PA 19104–6241. Email:
et al., 2000). Land may also be considered part of the
rozin@psych.upenn.edu.
person’s (or nation’s) extended self, and hence have a
325

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
326
particularly close relation to the person (or nation) (Belk,
positive contagion may trump negative contagion; at least
1988). Finally, land is perhaps the most important man-
there are many examples in the world of groups very at-
ifestation of the symbolic value of property in reinforc-
tached to land that they think was once theirs, but which
ing group identity (Ledgerwood, Liviatan & Carnevale,
has been occupied by the “enemy” for centuries. In this
2007).
study, we explore the relation between attachment to the
In this exploratory study, we develop a measure of at-
land of Israel and measures contagion. We also consider
tachment to national land (space) and sacred land (place),
some other aspects of land attachment, including its re-
in the context of the attachment to the land of Israel by Is-
lationship to political views about Israel, religiosity, the
raeli and American Jews. As we developed this measure,
personal importance of Israel, aversion to Arabs, views
we came to realize that the space/place distinction, may
on the vulnerability of Israel, and attachment to personal
be dif?cult to instantiate and differentiate. Although we
land.
believe it is worth attention as one analyzes relationships
We examine attachment to the “land of Israel” for Jew-
to land, in the present study we have employed a single
ish college students in both Israel and the United States.
composite measurement of attachment to the land of Is-
There have been important studies on the land disputes in
rael.
Israel-Palestine, particularly from the perspective of his-
The focus of this paper, in addition to measuring de-
tory and political science (e.g., Lustick, 1993), intergroup
gree of attachment to land and individual and cultural
attitudes (e.g., Bogardus, 1955) and acceptable and unac-
differences in this variable, is to explore one interesting
ceptable tradeoffs (Ginges et al., 2007). Land is central
potential determinant of land attachment, and that is con-
to this particular con?ict, with both sides considering the
tagion beliefs. Contagion beliefs are based on the sympa-
land to be “theirs.” In many cases, the contested land
thetic magical law of contagion: “once in contact, always
is predominantly populated by members of the opposing
in contact” (Frazer, 1922/1890; Mauss, 1972/1902; Ty-
group. Issues of claim over sacred land — that is, sites
lor, 1974/1871). When two objects are in physical con-
with biblical or historical signi?cance such as the Temple
tact, they pass “essences” between them, such that basic
Mount — come into con?ict as well.
properties of each are embedded in the other. This pro-
Personal land attachment is also relevant to and con-
cess, originally believed to be present only in the minds
founded with national land attachment in Israel-Palestine.
of people in traditional cultures, is now recognized as a
It is customary for Palestinians who were displaced to
universal part of adult thinking (Rozin, Millman & Ne-
wear the key from their former homes in order to remem-
meroff, 1986; Rozin & Nemeroff, 1990; Nemeroff &
ber their speci?c villages and the land on which they lived
Rozin, 2000). Contagion effects are typically permanent,
(Khalili 2005). There is evidence that for many Arabs and
and can be accomplished with minimal physical contact.
Jews, national land or particular sites in Israel constitute
The transfer of “essence” that psychologically character-
sacred, and hence protected values (Ginges et al., 2007).
izes contagion effects can be mentally represented in at
Although American Jews do not live in Israel, as part
least two ways (Nemeroff & Rozin, 1994). In one con-
of the Jewish Diaspora, they often have strong feelings
ception, the essence is material, and can be eliminated by
about the land of Israel. As a major force in the support of
physical processes. In the other, the essence is spiritual,
Israel (functioning like many other diaspora), understand-
and is resistant to erasure by physical processes.
ing the degree and nature of this type of land attachment
Negative contagion is much more common and more
is of relevance to understanding ethnopolitical con?ict.
potent than positive contagion (Rozin et al., 1989; Rozin
& Royzman, 2002). Many people do not show posi-
tive contagion, whereas everyone shows negative conta-
2 Method
gion. Furthermore, for individuals who experience both,
a combination of the two (e.g., a food touched by a cock-
Israelis were 187 psychology, sociology, business admin-
roach and one’s favorite personal memento, or a sweater
istration, and education students from the Hebrew Uni-
worn by Adolph Hitler and Mother Teresa) almost invari-
versity in Jerusalem, and psychology students from the
ably results in a strong negative outcome. When a group
Open University in Givat Ram, recruited as volunteers
has lived and died on a particular area of land for years,
from their classes. They received research credit for their
decades, or centuries, one can imagine that their essence
participation. We eliminated anyone with more than 25%
is passed into the land. This could form the base for a land
missing values (3 participants). The resulting sample size
attachment based on positive contagion. On the other
was 184. The average age was 25.1 years, ranging from
hand, if “one’s land” was occupied by the “enemy” for
12–58, and 66% of the sample was female. The Ameri-
some period of time (or if it was conquered from the “en-
can sample was 143 Jewish students from the University
emy”), there could be a strong negative contagion effect.
of Pennsylvania, reduced to 134 by the elimination crite-
It appears that land may be one of the rare cases where
ria. The American questionnaire was posted on a website,

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
327
where students could complete it as one option to ful?ll
0.1
a research participation requirement in introductory psy-
30
NAMECNTRY
0.09
chology. The study was restricted to students who had
ISR
USA
at least one Jewish parent. The sample was 51% female,
25
and the average age was 19.8 years, ranging from 18–32.
0.07
The questionnaire employed in the study was anony-
20
mous. It consisted of twelve demographic questions such
as gender, citizenship, age, and where the participant had
Count
15
0.05
lived since birth. We also assessed contagion sensitiv-
ity, religiosity, the importance of Israel, political views
Proportion per bar
about Israel, aversion to Arabs, perceived vulnerability
10
0.03
of Israel, and attachment to personal land. The particu-
lar items are described along with the results. The ques-
5
tionnaire was originally written in English and two sepa-
0.01
rate and independent Hebrew translations were made by
0
0
1
2
3
4
5
6
7
8
9
10
11
native speakers of both languages. The two translations
were compared and combined to form a ?nal Hebrew ver-
GRPLNDAVG7
sion. When the translations were discrepant, a third party,
Figure 1: Distribution of GROUPLAND7 by GROUP.
native speaker of both languages, resolved the difference.
The substantive questionnaire items are presented in the
results and tables.
ther (3) Syria or (4) Jordan (two separate items) for equiv-
alent land owned by Syria or Jordan, with the end of
3 Results
straightening out the border; and (5) willingness to make
this type of trade in the abstract (without specifying the
precise location of the land). The sixth item was a 108
A substantial proportion of participants endorse untrad-
mm line, anchored at the left by the words “easily trad-
ability of land. For example, 65% of Israelis and 86%
able, e.g. your couch” and at the other end by “never trad-
of Americans answer “No” to the question “Is there any
able, e.g., your child”: participants were asked to mark on
land anywhere for which you would trade a parcel of East
the line where the tradability of the land of Israel would
Jerusalem?” and 65% of Israelis and 53% of Americans
go on that line (the score was millimeters from the left
answer “No, I would not be willing to trade this” in ref-
point divided by 10). The seventh item was a summary
erence to the proposed trade of “A piece of land on the
willingness to trade score for the land of Har Hertzel (a
border of your country (referring to Israel) for a similar
cemetery in Jerusalem, where a number of individuals of
piece just across the border.” In response to the ques-
importance in the history of Israel are buried), averaged
tion “Is there any piece of land in Israel that you would
across three items (discussed below and converted to a 0
never be willing to trade under any circumstance?”, 59%
to 10 scale).
of Israelis who answered this question speci?cally nom-
These seven items might be sorted conceptually into
inated “Jerusalem” or the “old city.” In response to an
four involving national land (“space,” items 3–6 above),
item about the tradability of Har Hertzel, a cemetery in
and three about sacred land (“place,” items 1,2 and 7
Jerusalem that holds the remains of some major ?gures
above). The mean of the four national (space) items cor-
in the history of Israel, 83% if Israelis and 70% of Amer-
related .50 with the mean of the three sacred (place) items
icans agreed that they “would never trade it for other land
for the Israeli sample, and .42 for the American sample.
or anything else.” All of these responses suggest the ex-
A factor analysis of the seven items revealed a ?rst unro-
istence of protected values.
tated factor that accounted for 38% of the variance, with
the lowest of the seven items loading .52 on this factor.
3.1 The measurement of group land attach-
A varimax rotation revealed two factors, but they did not
ment: GROUPLAND7
sort on the space/place variable. These ?ndings justi?ed
making a single score, GROUPLAND7. We adopt the
Our measure of land attachment was seven questions
custom of following any averaged variable with the num-
about willingness to trade national or “sacred” land. The
ber of component items that contribute to it — based on
?rst ?ve, answered as yes (scored 0) or no (scored 10)
the average of the scores on each of these seven items
referred to (1) the Temple Mount or (2) “Is there any
(each one adjusted to be on a 0–10 scale). Thus, although
land anywhere for which you would trade a parcel of East
we see conceptual differences between national and sa-
Jerusalem?”; trading non populated Israeli land with ei-
cred land, the statistical analysis suggests combining the

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
328
Table 1: Major variables and predictors of attachment to group land.
Item
Israel Mean (s.d.) US Mean (s.d)
r (all Ss)1 r (Israel)1,2 r (USA)1,3
GROUPLAND7
7.02 (1.79)
7.08 (1.39)
PVD5
5.67 (1.56)
6.24 (1.42)**
.16***
.18
.14
NEGCON34
7.30 (2.95)
5.59 (2.66)**
.26***
.20
.43***
POSCON6
6.98 (1.70)
7.15 (1.54)
.31***
.29***
.36***
POSCONNOLND4 7.22 (2.05)
6.96 (1.84)
.26***
.22**
.34***
POLITICS5
5.43 (2.16)
4.94 (1.51)
.60***
.61***
.59***
POLINOLAND2
5.78 (2.41)
3.93 (2.03)*** .31***
.38***
.27**
ARABAVER3
4.68 (2.79)
3.80 (2.50)**
.31***
.38***
.20
RELIGIOS
2.18 (2.74)
4.73 (2.99)*** .28***
.28***
.33***
IMPISRAEL
8.77 (1.54)
7.44 (2.53)*** .37***
.45***
.40***
VULNER9
6.64 (1.46)
8.17 (1.64)*** .21***
.08
.46***
PERSONALLAND
A: TLANDFAM
6.20 (3.70)
7.65 (3.17)*** .35***
.36***
.38***
B: MMFAM5
7.62 (2.67)
7.35 (2.38)
.37***
.40***
.30**
** difference signi?cant at p<.01, two tailed; *** p<.001
1 Correlation between GROUPLAND7 and each of items below. Given the ns involved,
correlations of .23 or more are signi?cant at p<.01 (2-tailed), and .29 at p<.001
For combined data (N=268), r=.16, p<.01; r=.21, p<.001
2 N= 182–184, except 84 MMFAM and 123 NEGCON3
3 N=134 for USA, except 102 for MMFAM and 113 for NEGCON3
4 Only 123 Israelis and 113 Americans received a negative contagion score because we
excluded any subject who rated either Hitler or Cockroach contagion as positive.
5 Only 84 Israelis and102 Americans answered this question
measures of these two aspects of land. The Cronbach
tion yielded 6 factors, with a clear break in the Scree plot
alpha for this 7-item scale was .66. There was no signif-
after 3 factors. We forced a three factor solution, and re-
icant difference in GROUPLAND7 between Israelis and
covered three varimax rotated factors that mapped quite
Americans (Table 1). The variation in GROUPLAND7 is
well onto the concepts of: 1) concern about infection and
quite high for both samples (Table 1, Figure 1).
interpersonal contact, 2) positive contagion, and 3) nega-
tive contagion. We included the 14 of the 18 items which
loaded at least .44 on one of the factors.
3.2 Measuring contagion with respect
The ?rst factor includes only items derived from
to land and other entities (PVD5,
Faulkner et al.’s (2004) perceived vulnerability to disease
NEGCON3,
POSCON6,
POSCON-
scale. The ?ve items that loaded at least .44 on this fac-
NOLAND4)
tor, all answered on a 1 to 5 strongly disagree to strongly
agree scale, were (R = reverse scored): “It really bothers
There is no existing scale for contagion sensitivity. Eigh-
me when people sneeze without covering their mouths.
teen contagion sensitivity items were collected from
I don’t like to write with a pencil someone else has ob-
Schaller’s work on perceived vulnerability to disease
viously chewed on. I prefer to wash my hands pretty
(Faulkner, Schaller, Park & Duncan, 2004) and our own
soon after shaking someone’s hand. I dislike wearing
work on contagion (Rozin, Fallon & Mandell, 1994;
used clothes because you don’t know what the past per-
Rozin, Markwith & Nemeroff, 1986; Rozin et al., 1989)
son who wore it was like. My hands do not feel dirty after
with a few items created for this study. A principal com-
touching money (R ).” This factor score, PVD5, was the
ponent factor analysis of the correlation matrix for the
average of the scores of each of these items, adjusted to
combined Israeli-American sample with a varimax rota-
be on a 0 to 10 scale.

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
329
A second factor dealt with positive contagion. Three
and American samples are presented in Table 1. Corre-
of the six items were (rated on the 5 point agree-disagree
lations for GROUPLAND7 with PVD5 are low, and sig-
scale): “1. If land was owned in a single family for gen-
ni?cant only for the combined sample (Israel .18, USA
erations, there would always be something of their ances-
.14; combined groups .16). Correlations with NEGCON3
tors on that land, even after it has changed ownership. 2.
(note the smaller n for this variable because of eliminated
If I wear my great grandmother’s wedding ring, I am able
participants as described under method) were larger and
to keep some of her spirit with me. 3. A piece of land on
signi?cant for Americans and the combined groups (Is-
which my ancestors are buried contains something impor-
rael: .20; USA: ,43, combined .26,). The correlations
tant of them, whether spirit or something else.” A fourth
with POSCON6 were all signi?cant (Israel: .29,; USA:
item was,: “use the . . . . scale ( –100, worst imaginable
.36; combined .31). Positive contagion without Land
thing, to +100, best imaginable thing) to rate the follow-
(POSCONNOLAND4) was signi?cantly correlated with
ing scenarios. Imagine that sitting in a chair is rated as
GROUPLAND7 (Israel: .43; USA: .22; combined: .26).
zero. . . . Rate sitting in that same initial chair after you
Since, in the small contagion literature, negative conta-
were told that David Ben-Gurion (the founding president
gion effects are always more powerful than positive con-
of Israel) owned and sat in this chair in the 1940s.” The
tagion, the closeness of the positive and negative conta-
remaining two items were from a set of items on untrad-
gion correlations are notable.
ability (choices: willing to trade, uncertain, not willing to
trade) dealing with an original family photo album and a
great grandmother’s wedding ring. The combined vari-
able, POSCON6, was the average of the six scores, each
3.4 Other predictors of group land attach-
converted to a 0 to 10 scale, with 10 indicating the maxi-
ment
mal positive contagion score.
The third factor, negative contagion, included three
3.4.1 Political
attitudes
(POLITICS5,
POLI-
items, all in the same format as the Ben Gurion item de-
NOLAND2)
scribed above, and rated on the same –100 to +100 scale.
They were aversion to juice contacted by a cockroach [1]
Attitudes to the land of Israel are a major aspect of Israeli
or sterilized cockroach [2] and aversion to a chair that had
politics. The Israeli “right” considers the land of Israel as
been used and owned by Adolph Hitler [3]. The com-
non-negotiable as a political position. It is very dif?cult
bined variable, NEGCON3, was the average of the three
to separate and causally align land attachment and politi-
scores, each converted to a 0 to 10 scale (but see note 4
cal views. We measured political views in terms of agree-
of Table 1).
ment on the standard 5-point agree-disagree Likert scale
The positive contagion scale contains two items that
with ?ve statements: “All of ancient Judah and Samaria,
refer speci?cally to positive contagion about land. We
up to the Jordan river, must be part of Israel. Israel should
created a fourth measure, POSCONNOLAND4 without
respond each time there is a terrorist attack with a more
these two items, so that land measures would not appear
powerful attack against the Palestinians. Israel is a coun-
in both the land attachment and contagion variables.
try that should only be for Jews. In order to complete
The four scales created correlate between .04 and .36
a peace treaty, I would be willing to give up control of
with each other (except for POSCON6 and POSCON-
the Temple Mount and the Western Wall to neutral in-
NOLAND4, which have overlapping items), consider-
ternational control (reverse scored). Israel should remove
ing the set of pairs for the Israeli and American sam-
west bank and Gaza settlements (reverse scored).” We av-
ples separately. The correlations are suf?ciently low
eraged the scores on the ?ve items, calling it POLITICS5
that we continued to use the three separate factor mea-
(Cronbach ? = .62). There was no signi?cant difference
sures. Americans were signi?cantly higher on PVD5
between Israelis and Americans (Table 1).
(essentially, fear of infection), Israelis were signi?cantly
Of all the variables we examined, POLITICS5 cor-
higher on negative contagion (NEGCON3), and there
relates most highly with GROUPLAND7 (r=.61 for Is-
were no signi?cant country differences for positive con-
raelis, r=.59 for Americans, combined r=.60) (Table 1).
tagion (POSCON6) or positive contagion with land fac-
Two of the ?ve political items (on terrorism and on Israel
tored out (POSCONNOLAND4) (Table 1).
only for Jews) do not directly involve land. We averaged
just these two items as a measure of political views “inde-
3.3 Relation of contagion to attachment to
pendent” of attitudes to land, per se (POLINOLAND2).
Group Land
Israelis were signi?cantly higher on this item (Table 1),
and the correlations of this item with GROUPLAND7
The correlations between the four measures of conta-
were signi?cant (r=.38 for Israelis, r=.27 for Americans
gion and the group land measures for both the Israeli
and r= .31 for the combined groups (Table 1).

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
330
3.4.2 Aversion to Arabs (ARABAVER3)
of other countries’ intentions. I believe that my coun-
try must be constantly alert for possible danger. I believe
Negative attitudes (aversion) to Arabs could also be ex-
that other countries will try to deceive my country if given
pected to be related to land attachment. Arab aversion
the chance. I believe that the things most important to
was measured using the framework provided by Bogar-
my country are at risk. I believe my country generally
dus (1933), as modi?ed by our previous measure for the
should not trust other countries. I believe other countries
study of aversion to Germans by Jews (Cherfas et al.,
are often unfair to my country. I believe my country is
2006). Respondents rated agreement (on a standard 5
criticized by other countries more than it should be.”
point scale) with 3 items (ARABAVER3): “I would be
A factor analysis (varimax) indicated that these 9 items
uncomfortable living next door to a Muslim family from
could reasonably be represented in a single factor, so
Saudi Arabia; I would be uncomfortable living next door
we averaged the scores on these items to create VUL-
to a Palestinian family; I would be uncomfortable living
NER9 (Cronbach’s ? = .88). VULNER9 is rated signif-
next door to a Christian Arab man from Lebanon.”
icantly higher by Americans than by Israelis (Table 1).
Cronbach’s ? was .83. Israeli and American sam-
For the combined sample, VULNER9 correlates signif-
ples did not differ signi?cantly (Table 1). ARABAVER3
icantly .21 with GROUPLAND7, but strikingly less for
correlated signi?cantly (.31) with GROUPLAND7 for
the Israeli sample (r=.08) than for the American sample
the combined sample and for Israelis (r=.38), and non-
(r=.46). This difference is highly signi?cant (p<.001).
signi?cantly for Americans (r=.20). Not surprisingly,
VULNER9 correlates .20 (p<.001) with POLITICS5 for
POLITICS5 correlates highly with ARABAVER3 (.36
the combined sample (r=.24 [p<.01] for Israelis and r=.37
for USA, .41 for Israel, .40 for combined sample).
[p<.001] for Americans).
3.4.3 Religiosity and importance of Israel (RELI-
GIOS, IMPISRAEL)
3.5 Assessment of predictors of GROUP-
LAND7
We employed two items to assess identity: “1. How re-
ligious are you? (0 = not at all, 1 = slightly, 2 = moder-
We have identi?ed eight variables (Table 1) that are
ately, 3 = extremely); 2. “How important to you is Israel”
candidates for explaining GROUPLAND7.
These
(rated on a 7-point scale from 1= not at all important to
are:
PVD5, POSCON6, NEGCON3, POLITICS5,
7 = extremely important). These two items correlated .52
ARABAVER5, RELIGIOS, IMPISRAEL and VUL-
for Americans, but only .09 for Israelis. For this reason,
NER9.
we did not combine the items, and instead treated each
A straight linear regression of the eight predictor vari-
separately.
ables on GROUPLAND7 was carried out for the entire
RELIGIOS is much higher for Americans, while IMP-
sample. (The n for this sample was only 235 [as op-
ISRAEL is higher for Israelis (Table 1). RELIGIOS cor-
posed to the full n of 318] primarily because we did not
relates signi?cantly with GROUPLAND7, .28 for Israelis
have NEGCON3 for many subjects, because of positive
and .33 for Americans, as does IMPISRAEL (.45 for Is-
scores on Hitler or cockroaches). The analysis yielded
raelis, .40 for Americans) (Table 1). Not surprisingly
a squared multiple R=.46. Signi?cant beta values, in
RELGIOS correlates with POLITICS5 (.42 for Ameri-
decreasing order, are POLITIC5 (beta =.459; P<.001),
cans, .38 for Israelis) and IMPISRAEL also shows sub-
IMPISRAEL (.159, P<.01) and POSCON6 (.107, P<.05).
stantial correlations with POLITICS5 (.33 for Americans,
No other variable showed an effect signi?cant at p<.05.
.38 for Israelis).
The relative independence of the in?uence of politics
and contagion is re?ected in the fact that the correla-
tion, for the full sample, between POLITICS5 and NEG-
3.4.4 Danger/vulnerability/distrust (VULNER9)
CON3 is .20 (p<.01), and POLITICS5 and POSCON6
We included 9 items to measure the degree to which par-
is .12 (n.s.). In contrast to contagion, some of the other
ticipants felt that Israel was vulnerable and felt distrust
variables correlated substantially with POLITICS5, and
towards other countries with respect to their support of
hence did not emerge as signi?cant in the regression (e.g.,
Israel. Six of these items are modi?cations of two core
ARABAVER3: r=.40).
beliefs (vulnerability and distrust) described by Eidelson
The results of any regression are, of course, a function
and Eidelson (2003). The items, all rated on the standard
of the variables entered. We now consider three reason-
5-point agreement scale, all referred to Israel. The items
able modi?cations of this regression. First, we eliminated
are: “My country has been the victim of force and vio-
NEGCON3, because it reduced the n from 315 to 236. In
lence by other nations. My country’s right to its home-
this regression, which still had a squared multiple r of .46,
land has not been adequately recognized by the rest of
POSCON6 emerged as the second best predictor, after
the world. I believe my country should be suspicious
POLITIC5. POSCON6 showed a beta of .219 (p<.001).

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
331
Overall, the most reasonable regression might be one
grew up, your family’s land, etc).” (The tradability line
in which individual items that directly dealt with land
was anchored at the left by the words “easily tradable,
were eliminated as predictors.
To accomplish this,
e.g. your couch” and at the other end by “never tradable,
we carried out a regression using 7 predictors (not in-
e.g., your child”).The score was millimeters from the left
cluding NEGCON3), but used as the political variable
point divided by 10. Participants were instructed not to
POLINOLAND2, composed of two political beliefs nei-
mark this line if there was no personal land that quali-
ther of which dealt directly with land.
Correspond-
?ed; 84 (47%) of Israelis and 106 (79%) of Americans
ingly, we left out the 2 of 6 positive contagion items
completed this item.
that dealt with positive contagion in land (the vari-
The two personal items correlated .53 for the Israelis
able POSCONNOLAND4).
For the combined sam-
and .14 for the Americans. Given the low correlation
ple, in this regression (R2 =.31), RELIGIOS had the
for Americans, and the large difference in number com-
highest beta (.211, p<.001), with POSCONNOLAND4
pleting the two personal land items, we performed sep-
slightly behind (.205, p<.001). With this set of variables,
arate analyses with each item. As indicated in Table 1,
for the USA sample alone (R2=.32), there were only
the A item correlated .36 (p<.001)with GROUPLAND7
two major predictors: VULNER9 (beta=.281, p<.001)
for Israelis, and .38 (p<.001) for Americans. Ameri-
and POSCONNOLAND4 (.231, p<.001). For Israelis
cans showed a higher degree of attachment to this per-
(R2=.41), the four best predictors (all signi?cant at p<.01
sonal land (mean=7.62) than Israelis (mean=6.00; table 1,
or better) were, in order, IMPISRAEL (beta = .288),
t(268) = 3.856, p<.001). The relation between item A and
RELIGIOS (.227), ARABAVER3 (.207), and POSCON-
GROUPLAND7 was not entirely mediated by religiosity
NOLAND (.193). There is evidence in these regressions
and politics, since when we regressed GROUPLAND7
that positive contagion without land in its measurement
against PERSONAL LAND A, RELIGIOS, and POLIT5
remains a predictor of land attachment, independently of
for Israelis and Americans separately, in both cases the
other major predictors.
contribution of item A remained signi?cant.
The B item, with a much smaller number of respon-
3.6 Relation of group land and personal
dents, rated on the 108 mm line, did not show a signif-
icant difference between Israelis and Americans (Table
land attachments
1). Personal land B correlated .40 with GROUPLAND7
A link between group and personal land would support
for Israelis and .30 for Americans. We conclude that
either of two hypotheses: (1) personal land may be the
there is a signi?cant and moderately large relationship
evolutionary foundation for feelings about group land, or
between personal land and group land attachment, sup-
(2) The culturally derived concept of property lies behind
porting the evolutionary hypothesis. However we recog-
both types of land attachment. Our measures of personal
nize that these ?ndings are very tentative, since personal
land attachment in this study were quite preliminary, and
owned land was not a relevant measure for about half of
compromised by the idea that personal land ownership
the Israeli students.
can be conceived as occurring over much longer time pe-
riods (biblical) or much shorter periods (founding of the
3.7 Israeli vs. American differences
state of Israel) for Israelis than for Americans. Home and
personal land ownership is probably considerably higher
For the mean scores on the 14 variables in this study,
in Americans. Also, college students are not the ideal
there are signi?cant differences (p<.01 by two group t-
informants on the subject of personal land.
tests) in eight cases (displayed with signi?cance levels in
We included two items that measured attachment to
Table 1). Notably, the groups do not differ in GROUP-
personal land. One (A) employed a hypothetical situa-
LAND7. Israelis are signi?cantly less sensitive to infec-
tion, using the scale: “Yes, I would be willing to trade
tion contagion, but signi?cantly more sensitive to nega-
this (coded as 0); Uncertain; I would need more details
tive contagion. Americans show more attachment to per-
about the trade (coded as 5); No, I would not be will-
sonal land (A version of question), and Americans report
ing to trade this (coded as 10)” as responses to “Your
Israel as much more vulnerable than do Israelis (Table
family-owned piece of land that has belonged to your
1). Israelis rate the importance of Israel as signi?cantly
family since the establishment of the state of Israel for
higher, but Americans score much higher on religiosity.
a somewhat larger and better piece of land nearby.” The
The difference between the groups on religiosity is the
second item (B) was parallel to the sixth item described
largest difference between the groups. RELIGIOS and
above for GROUPLAND7, and employed the same 108
IMPISRAEL are closely linked for Americans (r=.52),
mm long analog scale. Participants were asked to mark
and much (and signi?cantly) less so for Israelis (r=.09).
on the line the tradability of “the piece of land that is of
The only signi?cant difference between the groups in cor-
most personal value to you (e.g. the land on which you
relations with GROUPLAND7 is the higher correlation

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
332
all respondents, there are signi?cant drops in untradabil-
Table 2: Tradability of site of Har Hertzel, percent re-
ity with the earthquake, and again with the prison (?2
sponding.
4,
p<.001 in all cases). A minority of individuals holds
Israel Israel Israel
USA
USA
USA
to untradability even after the land is “contaminated” by
will-
possi- unwill- will-
possi- unwill-
presence of “the enemy.” For most respondents, it is the
ing
ble
ing to ing
ble
ing to
topsoil, which contains the remains of some major ?g-
trade
trade
ures in the history of Israel, which is the only untradable
entity.
Hertzel
2
13
85
5
26
68
Quake
19
42
39
14
60
26
Prison
36
36
28
33
54
13
4 Discussion
N=178 Israel, 133 USA; ?2 is signi?cant at p<.001 for
both transitions (Hertzel to Quake and Quake to Prison)
We have developed a measure of attachment to land, and
for both Israelis and Americans.
?nd that for many Israelis and Americans, land is an un-
tradable entity, a protected value. The tradability of Is-
raeli land does not differ between Israelis and American.
between vulnerability and GROUPLAND7 for Ameri-
We have presented preliminary evidence that attachment
cans (.46) than Israelis (.08). In general, for Americans,
to personal land is related to tradability of group land.
vulnerability is closely linked to RELIGIOS (r=.40) and
One possible interpretation of this link has evolutionary
IMPISRAEL (r=.59). The equivalent values for Israel,
implications, insofar as territory, a feature of nonhuman
both .15. suggest a much weaker linkage for Israelis, as
animals, may be preadapted for attachment to land rep-
well as a lower sense of vulnerability.
resented as something more than a source of resources.
Alternatively, attachment to land may be a manifestation
of the concept of property, which applies to a wide range
3.8 Indications of the nature of land conta-
of human activities, and may or may not be linked to kin
gion
or territory in non-human primates.
Our most interesting ?nding is that both positive and
To explore what aspects of land motivate an attachment
negative contagion beliefs are related to land attachment.
to it, participants were asked to use the following scale to
Two of the four items in the positive contagion scale re-
answer three questions: “Glad to trade it for equivalent
fer speci?cally to positive contagion transmitted through
land (coded as 0); might consider trading it for land of a
land, by ownership or burial of ancestors in the land.
greater monetary value (coded as 5); would never trade
The other four items refer to contagion through a ring,
it for other land or anything else (coded as 10).”
family album, or contact with David Ben Gurion. Cor-
“1. The tradability of Har Hertzel (a cemetery in
relations with group land are nonsigni?cantly higher for
Jerusalem, wherein are buried many major ?gures from
the POSCON6 (.31) than for POSCONNOLAND4 (.26),
the history of Israel, including Theordore Hertzel and
suggesting that, although there may be some speci?c con-
Prime Minister Itzak Rabin).”
tribution of positive land contagion, some general pos-
“2. Now imagine the following: There has been an
itive contagion sensitivity is also involved. The results
earthquake in Israel. All of the graves on Har Hertzel
on Har Hertzel suggest that land attachment and positive
have been uncovered and the ?rst 50 feet of topsoil on the
contagion effects are principally resident in the top 50
mountain has been destroyed. The government decides
feet of the land, but that for some, the attachment may
to move all those who were buried on Har Hertzel to a
go deeper, and even survive negative contagion effects.
different burial site. Tradability of site of Har Hertzel.”
Although land attachment is clearly related to political
“3. After the earthquake and the removal of the graves,
views and Jewish religiosity or importance of Israel, we
a prison is built on Har Hertzel to hold Palestinians con-
present evidence that the positive contagion contribution
victed of violent political crimes. The prison is built and
to land attachment cannot be entirely mediated by these
has been established in this location for ten years. Trad-
other beliefs and attitudes.
ability of site of Har Hertzel.”
Our analysis reveals an interesting contradiction in our
Israelis are signi?cantly more reluctant to trade this
data, and in the real world. Land that has been inhabited
land, in its present form, or after degradation, than Amer-
primarily by one’s “enemies” (such as, from the Israeli
icans (Table 2). While Har Hertzel itself was untradable
point of view, Gaza or parts of the West Bank, or in our
for 85% of Israelis, this drops to 39% after the earth-
Har Hertzel example, the land which for ten years housed
quake, and 28% after the prison. Equivalent values for
a prison for Palestinians) should become more negative
Americans are 68%, 26% and 13%, respectively. Across
to those more sensitive to negative contagion. Negative

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
333
and positive contagion sensitivity are modestly positively
?uence of Diasporas in ethnopolitical con?ict and their
correlated (r=.30 in this study, for the combined sample).
approaches to con?ict resolution.
Our previous work on contagion (e.g., Rozin, Nemeroff
This study is merely a ?rst step. We examined only
& Millman, 1986; Rozin et al., 1989; Nemeroff & Rozin,
one locale, Israel-Palestine, and only college students. It
1994; summarized in Rozin and Nemeroff 1990 and Ne-
is possible that some respondents answered some ques-
meroff & Rozin, 2000), clearly indicates that negative
tions to support their political beliefs, since in Israel the
contagion dominates positive contagion. Yet many Jews
issue of the land of Israel is central in politics. Fur-
(and many of our Har Hertzel prison scenario respon-
ther research would be facilitated by development of re-
dents) maintain a strong attachment to enemy-occupied
liable and valid positive and negative contagion sensitiv-
land. One possible account for this, not previously stud-
ity scales. Follow-up studies should include older adults,
ied, is that, if certain things are initially positive (e.g.,
and both Arabs and Jews (see Ginges et al., 2007, for
originally possessed by the self or an admired other), sub-
an excellent example of selection of relevant samples).
sequent negative contacts are less contaminating. That
Similar explorations would be desirable in other cases of
is, there may be a priority principle in contagion. None
land contestation, including Iraq, Northern Ireland, and
of our examples in our previous studies involved a clear
Sri Lanka.
long-term temporal sequence, in which the positive con-
tact was for a long period and preceded the negative con-
tact. Thus, we know a sweater worn by both Hitler and
References
Mother Teresa would be rated as negative, but we don’t
know that a sweater worn and owned by Mother Teresa
Baron, J., & Spranca M. (1997). Protected Values. Orga-
for years would be rendered negative if Adolph Hitler
nizational Behavior and Human Decision Processes,
later wore it once. For the present case of the land of
70, 1–16.
Israel, land believed to be originally “yours” is relatively
Belk, R. W. (1988) Possessions and the Extended Self.
resistant to negative contagion effects. Furthermore, in
Journal of Consumer Research, 15, 139–168.
the general domain of “property,” at least for Western-
Bogardus, E. (1933). A Social distance scale. Sociology
developed cultures, it is our sense that a valued entity that
and Social Research, 17, 265–271.
belonged to a person and then illegitimately becomes the
Bogardus, E. S. (1955) Reducing Social Distance be-
possession of another, remains the valued and legitimate
tween Arabs and Jews. Sociology and Social Research,
property of the original owner. This issue, with respect to
40, 41–48.
land and other forms of property, requires further empiri-
Deudney, D. (1996). Ground identity: Nature, place, and
cal study.
space in nationalism. In: Y. Lapid & F. Kratchowil
Our measure of contagion (negative or positive) is not a
(Eds.). The return of culture and identity in IR theory,
valid scale — there is none, at this time — and further re-
pp. 129–145. Boulder, Colorado: Lynne Rienner.
?nement might facilitate further investigation. As it turns
Faulkner, J., Schaller, M., Park, J. H., & Duncan, L. A.
out, negative contagion sensitivity, operating in terms of
(2004). Evolved disease-avoidance mechanisms and
occupancy of land by the “enemy,” should encourage re-
contemporary xenophobic attitudes. Group Processes
ductions in land attachment. It may be that those who
and Intergroup Relations, 7, 333–353.
show a negative contagion effect from land occupied by
Fiske, A., & Tetlock. (1997). Taboo tradeoffs: Reactions
the “enemy” may also show more positive contagion with
to transactions that transgress spheres of justice. Polit-
reference to prior ownership, as would be expected since
ical Psychology, 18, 255–297.
positive and negative contagion sensitivity are positively
Frazer, J. G. (1922). The golden bough: A study in magic
correlated. However, the underlying positive contagion
and religion (abridged ed., T.H. Gaster, Ed.). New
belief, based on origin, may be more important, in the
York: Macmillan. (Original work published 1890)
case of land, than the direct negative contagion effect.
Ginges, J., Atran, S., Medin, D., & Shikaki, K. (2007).
Our results on Jewish Americans suggest about the
Sacred bounds on rational resolution of violent politi-
same amount of attachment to the land of Israel as for
cal con?ict. Proceedings of the National Academy of
Israelis (Table 1). On the other hand, American Jews see
Sciences 104, 7357–7360
Israel as substantially more vulnerable than do Israelis,
Khalili, L. (2005). Commemorating contested lands.
and while American judgments of vulnerability correlate
In Lesch, A. & Lustick, I. (Eds), Exile and Return
substantially and positively with land attachment (r=.46),
(pp. 19–40). Philadelphia: University of Pennsylvania
the correlation for Israelis between perceived vulnera-
Press.
bility and land attachment is near zero (r=.08). Israeli-
Ledgerwood, A., Liviatan, I., & Carnebale, P. J. (2007).
American similarities and differences, after further ex-
Group-identity completion and the symbolic value of
ploration, may help inform us about the important in-
property. Psychological Science, 18, 873–878.

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Attachment to the land of Israel
334
Lustick, I. (1993). Disputed Territories, Unsettled Lands.
Rozin, P., Nemeroff, C., Wane, M., & Sherrod, A. (1989).
London: Cornell University Press.
Operation of the sympathetic magical law of contagion
Mauss, M. (1972). A general theory of magic (R. Brain,
in interpersonal attitudes among Americans. Bulletin
Trans.). New York: Norton. (Original work published
of the Psychonomic Society, 27, 367–370.
1902)
Rozin, P., & Nemeroff, C. J. (1990). The laws of sym-
Mayr, E. (1960). The emergence of evolutionary novel-
pathetic magic: A psychological analysis of similarity
ties. In S. Tax (Ed.), Evolution after Darwin: Vol. 1.
and contagion. In J. Stigler, G. Herdt & R. A. Shweder
The evolution of life (pp. 349–380). Chicago, IL: Uni-
(Eds.), Cultural Psychology: Essays on comparative
versity of Chicago Press.
human development (pp. 205–232). Cambridge, Eng-
Nemeroff, C., & Rozin, P. (1994). The contagion concept
land: Cambridge University Press.
in adult thinking in the United States: Transmission
Rozin, P., & Royzman, E. (2001). Negativity bias, nega-
of germs and interpersonal in?uence. Ethos, 22, 158–
tivity dominance, and contagion. Personality and So-
186.
cial Psychology Review, 5, 296–320.
Nemeroff, C., & Rozin, P. (2000). The makings of the
Tetlock, P. E., Kristel, O., Elson, B., Green, M., & Lerner,
magical mind. In K. S. Rosengren, C. N. Johnson, &
J. (2000). The psychology of the unthinkable: Taboo
P. L. Harris (Eds.). Imagining the impossible: magical,
trade-offs, forbidden base rates, and heretical counter-
scienti?c, and religious thinking in children (pp. 1–34).
factuals. Journal of Personality and Social Psychology,
New York: Cambridge University Press.
78, 853–870.
Rozin, P. (1999). Preadaptation and the puzzles and prop-
Tylor, E. B. (1974). Primitive culture: Researches into
erties of pleasure. In D. Kahneman, E. Diener & N.
the development of mythology, philosophy, religion, art
Schwarz (eds.). Well being: The foundations of he-
and custom. New York: Gordon Press. (Original work
donic psychology (pp. 109–133). New York: Russell
published 1871).
Sage.
Weber, E. (1977). Roads, roads and roads. In E. We-
Rozin, P., Millman, L., & Nemeroff, C. (1986). Opera-
ber From Peasants into Frenchman (Ch. 12). London:
tion of the laws of sympathetic magic in disgust and
Chatto and Windus.
other domains. Journal of Personality and Social Psy-
chology, 50, 703–712.

---