This is an open access work published under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0), which allows reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use (https://creativecommons.org/licenses/by-sa/4.0/)

1. Introduction

Since the day Franklin Roosevelt met Winston Churchill on U.S.S. Augusta to map out the Atlantic Charter, the cornerstone of the United Nations, many U.S. presidents have contributed to crafting and maintaining the “liberal international order.” Harry Truman founded the North Atlantic Treaty Organization; Lyndon Johnson signed the Treaty on Non-Proliferation of Nuclear Weapons; Richard Nixon and Gerald Ford started and concluded the Strategic Arms Limitation Talks with the Soviet Union and also joined the negotiations leading to the signing of the Helsinki Accords (Helsinki Commission Staff, 2019, pp. 1-2); Jimmy Carter signed several controversial human rights treaties; and Ronald Reagan celebrated the triumph of the western “liberal” front over the communist one.

If we define “order” as an “organized group of international institutions that help govern the interactions among the member states” (Mearsheimer, 2019, p. 9), “liberalizing” the international order proceeded with even more strength after the end of the Cold War. From the expansion of NATO and conclusion of the North American Free Trade Agreement during George H. W. Bush’s term in office to the signing of many international agreements, including the Kyoto Protocol to the U.N. Framework Convention on Climate Change, Marrakesh Agreement establishing the World Trade Organization, Comprehensive Nuclear-Test-Ban Treaty, and Rome Statute of the International Criminal Court by Bill Clinton, making a deal with more than 170 countries through the Paris Agreement on climate change, and concluding the long negotiations on the huge trade deal of the Trans-Pacific Partnership during Barack Obama’s term all clearly show the commitment of post-Cold War U.S. presidents to the liberal international order. Nevertheless, we should consider the foreign policy approach of the first years of George W. Bush’s presidency as an exception as it did not last long (Jervis, 2018, p. 4).

Since his inauguration in January 2017, Donald Trump devoted himself to rigorously questioning the general post-WWII U.S. foreign policy traditions and revising or abandoning them whenever was possible. It seems that he especially insisted on reversing Obama’s policies. The new president almost immediately pulled the U.S. out of the Trans-Pacific Partnership (TPP), the “world’s largest trade deal.” Began by G. W. Bush administration and after eight years of negotiations, TPP had been finally signed by Obama less than a year before Trump’s withdrawal (McBride, Chatzky & Siripurapu, 2021). Although some observers have praised the deal as “the most important U.S. geoeconomic response to the increasingly coercive weight of the Chinese economy in Asia” (Blackwill, 2019, p. 10), Trump called it “the job-killing disaster” (Trump, 2017c). Furthermore, trying to “fix” what Obama had done, Trump ordered the U.S. withdrawal from the Paris Climate Agreement and the Joint Comprehensive Plan of Action (JCPOA), two “noteworthy accomplishments” of Obama’s foreign policy (Boyle, 2017, p. 10).

That said, Trump’s opposition to international institutions was not limited to those of the Obama era. He additionally withdrew the U.S. from numerous multilateral agreements and international organizations one by one, including but not limited to the North American Free Trade Agreement (NAFTA), the Open Skies Treaty, the optional protocol to the Vienna Convention on Diplomatic Relations, the U.N. Human Rights Council, the World Health Organization (WHO), and the United Nations Educational and Scientific Organization (UNESCO)[1].

Moreover, Trump’s rhetoric clearly indicates his antagonism toward old traditions of conducting U.S. foreign policy. Some have mentioned his “persistent attacks on the alliance and on individual allies, particularly about burden-sharing but also on trade and other issues” (Sloan, 2020, p. 39). He called NATO “obsolete” and subsequently declared that he was “absolutely prepared” to leave “wealthy” countries to defend themselves (Powaski, 2019, p. 239). As an observer has duly noted, “Europe has had many fights with American Presidents over the years, but never in the seven decades… confronted one so openly hostile to its core institutions” (Glasser, 2018). On the other hand, Trump was unprecedentedly soft on some of the U.S.’ traditional foes, such as Russian and North Korean leaders. In fact, just at the same time that he called the E.U. a foe, he referred to Putin as “not an enemy” of his, and wished that “someday, maybe he’ll be a friend” (Trump, 2018b). Furthermore, the U.S. president expressed his positive attitude toward the North Korean leader, particularly by describing him as a “talented,” “very smart,” and trustworthy person (Trump, 2018c). All in all, it seems that, in the words of a reporter, Trump was “treating… historic friends as enemies and… historic enemies as friends” (The President's News Conference on Sentosa Island, Singapore, 2018).

Moreover, Trump apparently had no hesitation in attempting to solve foreign policy problems in eccentric ways. In hope of finally realizing the denuclearization of the Korean peninsula, for instance, Trump “unexpectedly” decided to be the first sitting president who met with a North Korean leader and talks to him directly, once in May 2018 and again in February 2019, despite domestic criticisms about “legitimizing” Kim Jong-un (van Buren, 2018). Another illustrative example of Trump’s unusual policy was when in December 2017, he recognized Jerusalem as the capital of the Zionist regime and ordered the U.S. embassy to be transferred from Tel Aviv to Jerusalem, a measure that faced global opposition at the time and was in sharp contrast with the policy that the United States and “virtually every other nation in the world” had pursued for decades. In fact, despite the Jerusalem Embassy Act of 1995 which mandates the relocation of the U.S. embassy, Trump’s predecessors, from Clinton to Obama, exercised the authority granted to them by the same act to postpone the mandate for consecutive six-month periods until June 2017 (Lander, 2017).

What is clearly prominent in the examples mentioned so far is “the unorthodoxy” of Trump’s foreign policy decisions in the eyes of the observers. Indeed, other U.S. presidents also took measures that could be called “unorthodox,” but considering the big picture of their foreign policy, those measures were relatively rare. Examples include Nixon’s unprecedentedly visiting China in 1972, Reagan’s unusual rhetoric in calling the Soviet Union “an evil empire,” Bush’s highly controversial war on Iraq in 2003 and his pulling the U.S. out of several international agreements, and Obama’s unprecedented communication with the Iranian president. However, although these examples resemble some of Trump’s deeds, it seems that no other post-WWII U.S. president has made so many unorthodox decisions throughout his tenure. Actually, we suggest here that Trump had an “unorthodox foreign policy” and by this, we mean not a few individual unconventional decisions, but an approach to foreign policy-making which generally is in sharp contrast with the long-established principles and practices. In this paper, we aim to explain this approach, and precisely, answer the following question: Why did Donald Trump pursue an unorthodox foreign policy, which was clearly in contrast with that of former U.S. presidents?

Even though representing Trump as a president who is surprisingly against the long-lasting liberal international order by his words and deeds is omnipresent, an explanation of his unorthodox foreign policy is relatively rare in the literature. Even one of the few studies (Clarke & Ricketts, 2017), which attempts to address “the causes of such a distinct break with recent American foreign policy,” simply puts Trump’s foreign policy along “the Jacksonian tradition” and therefore, fails to provide a satisfying explanation. Randall Schweller (2018a, 2018b) has extensively explained why he has found Trump “the right guy” who was pursuing the right policies according to the U.S. position in the international system but has not spelled out why his predecessors did not do so. Yet, another research paper (Abdiel, 2020), which focuses on Trump’s “irrational” decision regarding Jerusalem ultimately fails to answer the question we posed here due to the paper’s limited scope and some methodological shortcomings despite approaching a convincing explanation based on Trump’s leadership style. Similarly, Hassan and Featherstone (2020) who have used conceptual complexity to explain Trump’s “unpredictability doctrine” have not addressed the issue of his unorthodox foreign policy approach. Lastly, Noori and Hosseini (2019) who analyze the personality of Trump in order to show “the main element that directs” his foreign policy, do not address the significant difference between Trump’s foreign policy and that of his predecessors. By and large, it seems that the question we posed here remains unaddressed in the literature.

Our hypothesis is that Trump’s low level of cognitive complexity, compared with that of former U.S. presidents, is considerably responsible for his unorthodox foreign policy. In other words, as he perceives the world in a substantially different structure compared to his predecessors, his overall approach to foreign policy decision-making results in his markedly different foreign policy even in conditions where there have not been any substantial change in the structural setting. This is particularly the case when we compare Trump’s and Obama’s foreign policies as there was apparently no considerable sudden change in international or domestic factors during the presidential transition that could result in such an immediate and substantial revision of policies (recall the TPP withdrawal on Trump’s first Monday in the office).

In what follows, first, we will lay out the cognitive approach to foreign policy analysis, which attempts to explain foreign policy behaviors based on decision-makers’ cognitions with a particular concentration on cognitive complexity as the theoretical framework. Next, we will detail our method for measuring U.S. presidents’ cognitions, based on content analysis. In the third section, we test our hypothesis by comparing the results of content analysis for post-WWII U.S. presidents. Finally, we will conclude our argument by demonstrating how Trump’s outlying cognitive complexity might have considerable effects on pursuing an unorthodox foreign policy.

2. Cognitive Complexity in Foreign Policy Analysis

As we can think of a variety of factors that might affect foreign policy-making, various approaches to foreign policy analysis have developed as well. These approaches could be divided into two broad categories. While structural approaches tend to take international or domestic constraints as determinants of foreign policy behaviors, individualist approaches, without overlooking the structural factors, are based on the assumption that these factors “become related to the attitudes and decisions which, in the aggregate, comprise a state's foreign policy only by being apperceived and taken into account by those who participate in the policy-forming process” (Sprout & Sprout, 1957, p. 310). Here, either psychological factors (such as personality traits, emotions, and motivations) or individuals’ perceptions and misperceptions (Jervis, 2017) rooted in cognitions are taken to be most influential (see Cottam, Dietz-Uhler, Mastors & Preston, 2004, p. 20, 22; Young & Schafer, 1998) . Cognitions “include a wide variety of mental representations, schemas, models, categories, beliefs, values, and attitudes” (Winter, 2003, p. 116). We, as human beings, need to form cognitions in order to simplify our complex environment and make it comprehensible. How we think about ourselves, other people, the world around us, the interactions, the inferences, and so on, whether real or imaginary, existing or desired, constitutes our cognitions.

Although we do not tend to overlook the probable effects of non-cognitive factors on foreign policy decision-making, we suppose that incorporating them into the current research would be futile. As far as psychological factors are taken into consideration, they are just underlying factors that ultimately affect cognitions (Jervis, 2017, p. lxxiv). According to the nature of those factors, they are not only more difficult to measure but also apparently unnecessary to study as they only have an indirect influence on decision-making by affecting the process, and not the decision itself.

In fact, cognitions are very critical in foreign policy decision-making because:

The international environment imposes heavy information processing demands upon policy-makers... Policy-makers must deal with incomplete and unreliable information on the intentions and capabilities of other states. The range of response options is indeterminate. The probable consequences of each option are shrouded in uncertainty… Finally, to compound the difficulty of the task, policy-makers must sometimes work under intense stress and time pressure. (Tetlock & McGuire, 1986, pp. 149-150)

This is the very essence of “bounded rationality” (Simon, 1985, p. 294), which indicates that it is almost impossible to find a real-world decision-maker able to seek and choose the optimal option “objectively.” Instead, when a decision-maker deals with a situation, he or she begins to form his or her neither complete nor accurate perception of the situation by collecting information about the causes and consequences, the factors and people involved, the nature of these factors, and so on. This information would be received through some cognitive components, such as schemas, models, and beliefs. He or she, then, would specify and evaluate what can or cannot be done, and ultimately decide which option would better serve his or her goals based on cognitive components like beliefs, values, and attitudes. Thus, it seems that we “have no alternative but to explore in detail how each country's policymakers actually understand their situation” (Rose, 1998, p. 158).

The effect of decision-makers’ cognitions on foreign policy outcomes has been studied in numerous studies. Some of these address long-term strategies of multiple states in a region (e.g., Herrmann & Fischerkeller, 1995) and some focus on a government’s policy toward another one at a specific point in time (e.g., Ziv, 2011). The main decision-maker whose cognition has been taken to be most influential varies in different studies: heads of state/government (e.g., Schafer & Walker, 2006a), ministers (e.g., Walker & Falkowski, 1984), advisors (e.g., Walker & Schafer, 2000), central bankers (e.g., Thies, 2009), and so on. Furthermore, while some studies endeavor to contribute to theory development through observing large samples (e.g., Hermann & Hermann, 1989), others focus on studying a single or limited number of cases (e.g., Dyson, 2001).

Even though we used the general term “cognitions” when we briefly reviewed the above-mentioned studies, different cognitive components have in fact been examined there. Michael Young & Mark Schafer (1998) have classified the studies into four broad areas: operational codes, cognitive maps, images, and conceptual complexity. Operational codes are political belief systems that capture decision-makers’ views into “philosophical” (such as The “essential” nature of political life (i.e., harmonic or conflictual) and the predictability of the political future) and “instrumental” beliefs (such as the best approach for selecting political goals and the preferred ways of risk calculation and control) (George, 1969). On the other hand, beliefs regarding how the world works and how causes lead to consequences, that is, causal beliefs are organized in cognitive maps. Unlike operational codes, cognitive maps capture decision-makers’ beliefs on specific issues and depict them visually using directed graphs (Young & Schafer, 1998, p. 75). Yet, while operational codes and cognitive maps indicate decision-makers’ beliefs distinctively, images are overall perceptions of others as wholes (see Herrmann & Fischerkeller, 1995).

These three approaches can be more useful in explaining specific events in foreign policy rather than the overall foreign policy of an administration. Therefore, a different cognitive approach would be required, which we believe is cognitive complexity, a structural feature of cognitions. Though the term “cognitive complexity” is not usually used in the foreign policy analysis literature, we use it here as a higher-level term that encompasses both conceptual and integrative complexity. While now each of them has its own research program, they both have originated from the works of Harold Schroder and his colleagues on conceptual systems (Young & Schafer, 1998, p. 84). Conceptual complexity, however, is considered relatively stable over time, whereas integrative complexity has been defined as a situation-based feature (Suedfeld, Tetlock & Ramirez, 1977, p. 431).

Conceptual complexity mainly concerns differentiation. That is, “the degree of differentiation that an individual shows in characterizing other people, places, policies, ideas, or things” determines his or her level of conceptual complexity (Hermann & Hermann, 1989, p. 377). For example, one who sees all states as either “winners” or “losers” would be in a lower level of conceptual complexity compared to someone who believes in relative gains. Using a prevalent metaphor, conceptually complex individuals see “nuanced ‘shades of grey’,” while those who are conceptually simple see everything as either black or white (Dyson, 2009, p. 35). The results of Young & Hermann’s (2014) vast study on 123 leaders of the Pacific Rim indicates that a majority of them (i.e., 81 percent) during the period under study show little or no change in their conceptual complexity.

Studies on foreign policy often demonstrate the relations between conceptual complexity and some other stable patterns of behaviors. Some, for example, examine the relationship between the aggressive behavior of a state and its leader’s conceptual complexity (Hermann, 1980, p. 10). Some other studies suggest that low conceptual complexity corresponds to conflictual behaviors (Young & Schafer, 1998, p. 85). However, the influence of conceptual complexity on certain behaviors has also been studied. Foreign policy change, for example, is considered to be related to high levels of conceptual complexity (Yang, 2010).

Integrative complexity, on the other hand, focuses on situation-based cognitive complexity, as the individuals under different conditions might be at different levels of integrative complexity in various conditions. For example, the changes in leaders’ integrative complexity who were involved in the Cuban Missile Crisis (Guttieri, Wallace & Suedfeld, 1995), World War I (Suedfeld & Tetlock, 1977), and the Persian Gulf War (Wallace, Suedfeld & Thachuk, 1993) have been studied. Because of the situational feature of integrative complexity, we can not supposedly talk about an individual’s overall complexity. However, some researchers have measured the overall integrative complexity of 41 U.S. presidents, from George Washington to George W. Bush (Thoemmes & Conway III, 2007). In addition, the relationship between integrative complexity and personality traits, such as dovishness and hawkishness (Wallace, Suedfeld & Thachuk, 1993), being liberal or conservative (Conway III et al., 2016), and decision-making style (Suedfeld, 1994) has been studied. Such attempts raise a question in mind as to whether integrative complexity is relatively stable or it depends on the situation.

The utility of integrative complexity is in question in another aspect as well. In fact, one other difference between conceptual and integrative complexity is their respective constituents. Conceptual complexity only considers differentiation, but integrative complexity comprises both differentiation and integration. As Peter Suedfeld and Philip Tetlock (1977, p. 171) explain, “differentiation refers to the characteristics or dimensions of stimuli that are recognized and taken into account in decision-making… [while] [i]ntegration refers to the development of complex connections among the differentiated characteristics.” Individuals at the lowest level of integrative complexity have little ability to differentiate and no ability to integrate. The integration ability would rise only when the differentiation ability is at its highest level. Ultimately, at the highest level of integrative complexity, those two abilities reach their peaks. However, the results of the aforementioned study of 41 U.S. presidents indicate that none of these 41 leaders is even close to the level of high differentiation, and hence, none of them might have even a limited integration ability. Thus, it seems that focusing on differentiation solely would be enough.

Conceptual complexity has been adopted as the theoretical framework for the current research. The theorists and students of the conceptual complexity research program note some implications for high and low complexities, albeit only some of them have been empirically tested. Generally, high conceptual complexity corresponds with flexibility and showing interest in others’ opinions (Young & Hermann, 2014, p. 641). The opposite is true about the leaders with a low level of conceptual complexity:

Less complex leaders—with their lower cognitive need for extensive information search and examination of multiple policy perspectives—tended to be far less sensitive to both information and the external policy environment. This reduced sensitivity to information and to context manifested itself in limited information search and in limited emphasis upon the presentation by advisers of alternative viewpoints, discrepant information, and multiple policy options. Such leaders were more likely to rely upon simplistic analogies, “black-and-white” problem representations, or stereotypical images of their opponents during their policy deliberations. Further, given their limited interest in extensive policy debate or broad information search, low complexity leaders were also found to have, according to the archival evidence, very decisive and less deliberative decision-making styles. (Preston, 2001, p. 10)

Therefore, we expect that U.S. presidents, according to their respective levels of conceptual complexity, show the corresponding indications.

In addition to what we have mentioned so far, conceptual and integrative complexity have two very different methods of measurement. In the next section, we will briefly introduce them, but due to their incompetence, we will introduce our new method of measurement thereafter. Using the term cognitive complexity here also helps to methodologically distinguish our approach from the other two approaches.

3. Content Analysis

As it is impossible to take leaders to laboratories to evaluate their cognitions, analyzers of leaders’ cognitions have proposed “at-a-distance” measurement to overcome this challenge. This method is based on the analysis of the content of verbal materials, such as transcripts of speeches, interviews, news conferences, and letters (Winter, Hermann, Weintraub & Walker, 1991, p. 218). In fact, content analysis is a well-known technique for extracting some results, other than the apparent and explicit meaning, from specific content. More precisely, “[c]ontent analysis is a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use” (Krippendorff, 2004, p. 18). In this definition, the term “texts” means all sorts of communication media, including books, movies, messages, and so on. Content analyzers usually “count” and “record” certain parts or patterns in the contents and arrive at conclusions about them (Neuman, 2007, p. 227).

In the first place, the validity of using content analysis technique for assessing political leaders’ cognitions is based on the assumption that what is expressed by an individual represents his or her inner thoughts and feelings. Of course, the author (i.e., the political leader) can manipulate his or her audience by fabricating statements to communicate what he or she does not really think or feel. This flaw, unfortunately, may not be dismissed in general. However, under certain conditions, the assumption might hold true. First, if lots of materials that were addressed to different types of audiences and on different occasions were analyzed, the effect of “intentional deception” might be largely eliminated, since the same kind of deception could not be repeated everywhere (Schafer & Walker, 2006b, p. 47). Second, analyzing spontaneous materials (e.g., interviews or news conferences) would be preferable, since compared to prepared materials (e.g., speeches or letters), it is more likely that the political leader expresses what he or she really think, because of the short amount of time available to him or her to respond. Third, in the specific case of cognitive complexity, it is highly unlikely that the political leader could or even would want to mention all his or her remarks more complex or simpler than he or she thinks. In other words, “content is far more likely to be so manipulated than structural complexity” (Wallace & Suedfeld, 1988, p. 442), and hence measuring cognitive complexity by content analysis of lots of spontaneous verbal material may reasonably be considered valid and plausible.

Unlike clinical methods that are commonly used to assess psychological traits, content analysis is unobtrusive, that is, the subject is not aware of being under examination, and even if he or she knows about it, there is little chance that the kind of study or analysis would be known. This is particularly the case when we analyze spontaneous rather than prepared materials. Also, by relying on this kind of material, we can successfully eliminate the plausible effect of speechwriters or ghostwriters, since spontaneous statements originally belong to the political leader himself/herself, rather than to the speechwriters. That said, unobtrusiveness has its downside: researchers may fail to find what they seek. In fact, the individual whose verbal statements are being analyzed expresses what is supposed to be said and not what we look for, and hence, we may not find answers to our questions in his or her remarks. Nevertheless, in the case of political leaders, at-a-distance measurement seems to be our only option for assessing their cognitions.

As stated earlier, there are two very different methods for measuring conceptual and integrative complexity. However, the problem of indistinguishability in the results of both conceptual and integrative complexity studies and also the difficulty of coding the latter (see for example, Winter et al., 1991, p. 232; Thoemmes & Conway III, 2007, p. 204; Baker-Brown et al., 1992, p. 402) make both methods unsuitable.

As a result of the above-mentioned factors, it seems that we have no choice but to develop a new method for measuring cognitive complexity. Before we introduce it, we should note that by cognitive complexity, we mean “the ability to differentiate aspects or dimensions of the environment” (Winter et al., 1991, p. 222). We will use the Flesch-Kincaid readability score for measuring the complexity of texts. It is based on the assumption that the complexity of a text indicates the complexity of its author’s thoughts, and readability (or comprehensibility) could be a good indicator of text complexity. “Flesch-Kincaid readability score” has been developed to provide a yardstick for measuring the comprehensibility of U.S. Navy “training material,” by calculating the average words per sentence and average syllables per word and combining them in the following formula (Kincaid, Fishburne, Rogers, & Chissom, 1975, p. 14):

0.39 * (words/sentence) + 11.8 * (syllables/word) – 15.59

As the formula suggests, this score is based on two parameters: sentence length and word length (in terms of syllables). But how these two parameters are related to text complexity? G. McLaughlin (1969, p. 640) argues that these

are, respectively, indicators of semantic and syntactic sources of reading difficulty. In English, world length is associated with precise vocabulary, so a reader must usually make extra effort in order to identify the full meaning of a long word, simply because it is precise. Long sentences nearly always have complex grammatical structure, which is a strain on the reader's immediate memory because he has to retain several parts of each sentence before he can combine them into a meaningful whole. (McLaughlin,)

In fact, although it is theoretically possible that an author makes his or her sentences as long as he or she wants, it does not happen in practice. As a matter of fact, “[m]ost longer sentences… included noun modifiers, dependent clauses, nominalized verbs, deletions in coordinate clauses, appositives and clauses used as subjects” (Glazer, 1974, p. 467), which make them quite complex.

The result of Flesch-Kincaid scoring would be a school grade level required by a person to comprehend the text under evaluation. Lower scores (i.e., lower levels of education needed for comprehension) would indicate lower complexity of the text, and hence lower cognitive complexity of its author. It should be noted that we have found no evidence of deliberately expressing the thoughts in a simpler or more complex fashion than they actually are, particularly in spontaneous remarks. As a matter of fact, there seems no reasonable need to explain things in simpler language for the press, and on the other hand, the common practice of presidential news conferences demonstrates that answering a question with unnecessary complexity only brings more questions. Given this and also the spontaneous nature of responses in news conferences, we can confidently assume a direct relationship between the complexity of the presidents’ statements and the complexity of their thoughts.

4. Findings and Discussions

In order to compare Trump’s cognitive complexity to that of his predecessors, we collected the transcript of all news conferences held in the first year in office for each of the post-WWII U.S. presidents. We considered only the first-year statements because they were probably more representative of their author’s thoughts since the presidents presumably were less experienced in hiding them. These materials are freely available on The American Presidency Project website[2], which is maintained by the University of California at Santa Barbara. For Truman, Johnson, and Ford whose respective terms had not started on January 21, we considered 11 months and 10 days since their first day in office, equal to the first year of other presidents. Nevertheless, for each president, transcripts of solely their answers – and not their initial statements – in each news conference were obtained, and then, only those parts of the answers relevant to U.S. foreign policy were extracted in an attempt to eliminate the likely effect of issues on the complexity of thoughts. Also, among the extracted sentences, those with less than 5 words (e.g., “No comment,” “I do not know,” “Ask X”) were omitted, since it is highly unlikely that an idea could be expressed in such short sentences. Thereafter, these materials were aggregated per year, because, in contrast to speeches, news conferences do not have integrity. In other words, we might reasonably consider all news conferences held in a year as a giant news conference with numerous questions and answers. This is the case, particularly since some news conferences were too short, whereas some others were too long. Also, while some news conferences were concentrated on one particular issue, different kinds of questions were asked in others. Table 1 shows an overview of the examined materials. It should be noted that the varying number of news conferences is due to each president’s practice of holding presidential news conferences, and not because of sampling, since no sampling is performed. Nevertheless, we ultimately calculated the Flesch-Kincaid readability score for all statements expressed each year using the software developed by one of the authors and reported the results as the corresponding president’s cognitive complexity. Since the calculation was done by a computer software, the reliability of the measure is %100.

The results of measuring cognitive complexity for 13 post-WWII U.S. presidents are presented in Table 2. As we expected, Trump’s cognitive complexity is lower than all his 12 predecessors, and also his complexity is much below the average (by a z-score of -1.86). The greatest difference is noticeable between his and Obama’s complexity as Obama is the most cognitively complex president (by a z-score of +1.20), according to the results. It is noteworthy that the result of measuring Truman’s cognitive complexity might not be valid, since, unlike the other 12 presidents, his responses in his news conferences were usually very short, and also he did not elaborate on his beliefs or perceptions of the world very often. If we discard his cognitive complexity score, the average, the standard deviation, and the z-score of Trump’s cognitive complexity will be 10.56, 1.78, and -2.21 respectively. In other words, if we omit Truman’s score, we can more confidently infer that Trump’s cognitive complexity is an outlier.

These findings, particularly, are compatible with how observers assess Trump and Obama psychologically. Trump has been described as one who “oversimplif[ies] his world by viewing it with black-and-white thinking… or by reducing reality to an epic contest between righteous and evil, good and bad, winners and losers,” and also “unable to appreciate the complexity of either the external world or his internal life” (Frank, 2018, p. 14, 142). It has also been noted that “Trump stands out from other politicians as being very low in analytic thinking” (Jordan & Pennebaker, 2017, p. 312). Another observer suggests that “Trump’s speech is hardly bristling with complexity. Rather, his vocabulary is extremely simple, almost to the point of being childish” (Blair, 2016).

There are also reports that suggest Trump was not interested in taking advice from his advisors, which is compatible with what we expect from one with a low level of cognitive complexity. Regarding the second Trump-Kim summit in February 2019, for example, John Bolton, Trump’s national security advisor complains that “Trump wanted to meet Kim, and he didn’t want to hear anything contrary, which is probably why he didn’t want to hear me explaining that another meeting soon was a bad idea” (Bolton, 2020, p. 125). Another observer reports that “not only didn’t he read, he didn’t listen. He preferred to be the person talking. And he trusted his own expertise—no matter how paltry or irrelevant—more than anyone else’s” (Wolff, 2018, p. 114).

On the other hand, James Pennebaker, a psychologist who has examined the U.S. presidents’ remarks, has concluded that Obama “is the most complex thinker of the presidents… with John F. Kennedy as the runner-up. That's because Obama frequently presents an idea and re-evaluates it from a variety of perspectives [emphasis added] in subsequent sentences” (Landau, 2010). Another psychoanalyst has “accurately” attributed “complex… [and] nuanced thinking” to Obama (Frank, 2011, p. 236). Moreover, compatible with the implications of high cognitive complexity, there are numerous reports of Obama’s “good listening” and his openness to new information and advice (Décosterd, 2010, p. 4, 102; Kloppenberg, 2011, p. 74, 147; Frank, 2011, p. 99). Thus, the great difference that we measured here has been reflected in what has been reported on the two presidents as well.

But cognitive simplicity, indeed, is not equivalent to foolishness, insanity, or madness (Preston, 2001, p. 10). We absolutely do not suggest that Trump is a fool. Of course, he has had great achievements in his real estate business, and it is highly unlikely that an insane or a foolish person could reach such a position. All we argue here is that his worldview is, as Financial Times editor says, “transactional”, that is, to him what matters is winning or losing economically and tangibly (Stephens, 2017). In fact, if we take a look at different foreign policy issues in which Trump was involved, we could find much evidence lending credit to this assumption. His remarks on the 2016 Paris Agreement on climate change, for instance, is very illustrative:

[I]t sounds wonderful. [But, in fact,] [i]t's a disaster for this country. They've basically taken away your wealth, the Paris climate accord. And the other countries don't have to adhere to it. China doesn't kick in until 2030; they don't have to do anything until 2030. We had very high standards. We would have had to close, under some scenarios, 25 percent of our businesses in order to qualify under this ridiculous Paris climate accord. It sounds good. [But, actually,] [i]t was very bad and very expensive. (Trump, 2020)

Interestingly, in dividing the countries into winners and losers, he did not care much whether it is about the U.S. allies or adversaries. On NATO, for example, he believed that “NATO is wonderful, but it helps Europe more than it helps us. And why are we paying a vast majority of the costs?” (Trump, 2018d). Also, when asked to describe the U.S. relationship with Mexico, he responded that “as you know, Mexico—with the United States—has outnegotiated us and beat us to a pulp through our past leaders. They've made us look foolish” (Trump, 2017d). Regarding one of the traditional U.S. friends, Canada, he said that “[g]reat respect for Canada, great love for Canada, but it's been a one-sided deal for Canada” (Trump, 2017b). He stated the same things regarding the U.S. Asian allies:

I mean, I'll be honest, I just asked Japan. I said: “We're defending you. You're a very wealthy country. You're sending us millions of cars. You're making a fortune. We have a tremendous trade deficit with you. And we're defending you, and we're subsidizing your military with a massive amount of money”.

I said it to South Korea. We have 32,000 soldiers in South Korea. They're a very wealthy—these are great countries. These are very wealthy countries. I said, “Why aren't you reimbursing us for our costs?” And you know what? They look at me, and they can't even answer it, because there's no answer (Trump, 2018a).

This is similarly the case regarding the U.S. adversaries. On China, for instance, he believed that the main problem is that “for decades … it's been a very unfair trade situation” (Trump, 2017a). Also, he was concerned about the amount of money, “$1.8 billion in cash… tremendous money” (Trump, 2018e), when he asked about JCPOA. Another example is North Korea. Despite the criticism about offering concessions to Kim Jong-un by meeting him in person, Trump did not recognize any cost in participating in such a meeting and insisted that “I gave up nothing,” whereas he believed that in the past, “billions of dollars” were given to North Korea in vain (Trump, 2018c). As it is obvious in some of these cases, he thought that in foreign relations, the United States had been on the loser’s side of the deals before he came to office (“We don’t win anymore”) and he aimed to turn the page. For him, considering other perspectives were almost pointless.

On the contrary, it seems that other post-WWII U.S. presidents, more or less, tended to view the issues from more than one perspective. Crafting the liberal international order requires long-term thinking and sacrificing short-term benefits if necessary. The multilateral agreements that made this order possible usually incorporate giving concessions in terms of accepting limitations on freedom of action or ignoring some immediate gains, in hope of greater future gains that not necessarily economic. When Jimmy Carter, for example, was asked whether his emphasis on human rights “help or hurt” the people of the Soviet Union, he responded that:

It's hard for me to say. I think that in the long run our emphasis on human rights, the high publicity that has accrued to the human rights question because of the Helsinki agreement and the upcoming Belgrade conference in October--those two factors, combined, I think, dramatize every violation of human rights that is known.

And my guess is that the Soviets, like ourselves, want to put a good image forward for the world to observe, and I think in the long run that this emphasis on human rights will be beneficial to those who desire free speech and an enhancement of their own human freedoms (Carter, 1977).

Another example would be the 1972 Strategic Arms Limitation Talks (SALT) agreement, which was signed by Nixon, arguing that putting limitation on the U.S. and Soviet Union strategic weapons would be better than continuing the arms race and accepting the “the cost and… the danger of a [probable] nuclear confrontation” in the future (Nixon, 1970). Obama expressed his concerns that without international frameworks such as the 2016 Paris Agreement, which binds and limits the United States as well as other major polluting countries, “our kids will be choked off,” whereas “it's possible to grow the economy really fast and possible to bring down carbon emissions as well.” In fact, in sharp contrast with Trump, he believes that “international agreements, the tradition has been that you carry them forward across administrations… doing good for us [emphasis added] and binding other countries into behavior that will help us” (Obama, 2016).

This more or less “complex” thinking of Trump’s predecessors is not limited to strategic agreements. There are several instances of its application in other occasions of decision-making. When Dwight Eisenhower, for example, faces the decision of intervention in Indochina in 1954, he argues against the intervention, considering many perspectives, such as the posture of the U.S. “in the eyes of many Asiatic people,” difficulties of providing forces to replace French forces, and also “the attitude of the Congress and the people” regarding the probable intervention (Petersen, 1982, p. 1439). Additionally, when he was asked about his attitude regarding the St. Lawrence Seaway joint project with Canada, he responded that “There are so many controversial factors, and they seem to vary geographically as to their content, that I just think it takes a longer time than I have had to reach a real decision” (Eisenhower, 1953). As another example, when George H. W. Bush was defending NAFTA, he argued against the “siren’s call of protection” and in favor of open markets, because he believed that “[i]t's exports that have saved this economy as it goes through these tough times, and it is exports that will lead an extraordinary growth in the future” (Bush, 1992). We should also note the dissimilarities that Obama expressed between the situation in Libya and the situation in Syria when he had faced the decision of intervening in the latter in a National Security Council meeting:

Libya's fighting had taken place on an open desert, which allowed for clear targeting; Syria was enmeshed in urban warfare, with civilians, rebels, and soldiers intermingled. The Libyan rebels had had a chance at forming a cohesive government; there were no such possibilities in Syria. No other outside power was calling on the United States to intervene this time around. Finally, the conflict was cascading into a proxy war for the regionwide Sunni-Shiite confrontation. Not only did the United States have little at stake in this fight, but it also had little ability to influence its direction or outcome (Kaplan, 2016, p. 50).

5. Conclusion

The surprising or unorthodox nature of former U.S. president Donald Trump’s foreign policy is not a controversial fact. His opposition to the liberal international order and particularly the multilateral agreements, his unprecedented and unconventional decisions on several issues, and his unusual rhetoric toward U.S. allies and adversaries are indicators of this unorthodoxy. Since there are numerous examples of such behaviors, we can reasonably think of them not as a few exceptions, but as an approach. Trump’s approach to foreign policy is particularly in sharp contrast with that of his immediate predecessor, Barack Obama.

Attempting to answer the question as to why Trump’s foreign policy approach is so different from those of all post-WWII U.S. presidents in general, and Obama in particular, we hypothesized that cognitions, and more precisely, cognitive complexity is an important factor. Since the external and internal structural factors during the transition from Obama’s to Trump’s administrations are largely the same, we probably have to explore the individual-level factors, and cognitions play an important role in how individuals such as U.S. presidents make high-level decisions. Cognitions, actually, relate to how individuals perceive their environment and their options at hand. In fact, because the human mind cannot process all the information it gets, it has to simplify the reality through cognitive components such as beliefs, schemas, and images.

As we aimed to explain Trump’s overall approach to foreign policy and not a few specific decisions, studying the structure of the cognitions was considered to be more fruitful. According to the researchers, political leaders with a low level of cognitive complexity do not seek multiple perspectives on the situation and policy options and do not search for broader information but do rely on simplistic analogies and stereotypical images of their opponents. These characteristics lead to their decisiveness as well.

Measuring the cognitions of political leaders is usually done using the “at-a-distance” method, that is, by content analysis of leaders’ verbal statements. Two main at-a-distance methods for measuring cognitive complexity are frequently used, but according to the published results of the studies, both of them have produced relatively dissatisfying results. Thus, we decided to use the well-known Flesch-Kincaid readability score in order to measure the complexity of texts as an indicator of the complexity of their author’s thoughts. We collected the transcripts of all post-WWII U.S. presidential news conferences held in the president’s first year in office because their spontaneous nature could better reveal how these U.S. presidents really thought.

The results have demonstrated that, as we expected, Trump’s cognitive complexity is significantly lower than that of any other U.S. president since World War II, and mostly than that of Obama. Based on the implications of low-level cognitive complexity, Trump viewed the world mainly through a single perspective in which the U.S. in its foreign relations with the external world could be either winner or loser, in the economic and tangible sense of the word. He largely ignored other perspectives, whereas his predecessors, more or less, took into account other considerations, including U.S.’ soft power as we saw in the example of Eisenhower’s decision regarding intervention in Indochina, or long-term consequences of decisions, as it was the case in the example of Obama’s signature of the Paris agreement on climate change. Moreover, Trump’s refusal to search for more information or take advice likely led to making decisions based on his prejudice, which could result in deviation from the mainstream U.S. foreign policy since he had no prior experience in foreign affairs. In other words, listening to advisors who usually have considerable experience in foreign affairs could likely result in more conventional foreign policy decisions. Additionally, Trump’s low cognitive complexity left him with a few eccentric policy options, which he thought were rewarding despite the wide criticism. Other presidents, in contrast, due to their higher cognitive complexity, listened to their advisors, and hence, their decisions were generally along the orthodox path of U.S. foreign policy. Moreover, because of the decisiveness of cognitively simple leaders, it is less likely that Trump’s odd decisions would be back on track. By and large, we suggest that being at the lowest position of cognitive complexity among post-WWII U.S. presidents influenced Trump’s unorthodox foreign policy approach to a great extent.

It should be pointed out that whether Trump’s foreign policy approach was right or wrong, it was unorthodox. We did not aim to criticize this approach here, nor did we “cheer” for it. Our argument was that however international structural factors necessitated such an approach, the sharp contrast between Trump’s foreign policy approach and that of his predecessors needs an explanation, and we aimed here to provide such an explanation. Moreover, since there is not a definite resolution to the debate of whether “engagement with the world” is an appropriate approach despite the “declining” of the U.S. position internationally, it will be up to the U.S. presidents to decide for. As a matter of fact, the different policies that Trump and his predecessors pursued clearly indicates that structural factors cannot determine the foreign policy outcomes and the decision-makers should be studied as well.

While we solely pay attention here to Trump’s cognitions as the main factor that led to his unorthodox foreign policy, definitely there were also other individual-level factors that probably had some effects. One of these factors is his personality. His so-called “narcissism” and “low agreeableness”, for example, might exert some effects on the decision-making process—by designating officials who do not object to his decisions but simply implement them. Moreover, Trump’s emotions toward some persons might have affected a number of his decisions. Apparent hatred of Barack Obama might have led to Trump’s obsessive reconsideration of his predecessor’s policies. Furthermore, his seeming interest in Kim Jong-un’s personality might have played a role in Trump’s willingness to meet the North Korean leader. However, we believe that although these factors might explain some aspects of a number of Trump’s foreign policy decisions, they could not satisfyingly explain his unorthodox foreign policy approach, but rather complement the cognitive approach to make a more comprehensive explanation.

[1]. U.S. membership in some of these institutions, such as WHO and UNESCO, dated back to the 1940s.

[2]. https://www.presidency.ucsb.edu/