Six Connotations of Sustainability in Civil and Construction Engineering: A Corpus Linguistics Study

Abstract

This research explores the historic use of the concept of sustainability in civil engineering academic literature and identifies its conceptual evolution and connotative treatment over the last 40 years. We find that academic research addressing sustainability is exponentially increasing and that the focus on sustainability is enlarging from a primarily environmental perspective to include more social elements. Additionally, we propose a set of six connotations of sustainability: assessment, progress, permanence, abstractness, multidimensional, and intersectional. These connotations provide conceptual clarity for academics and practitioners in better-defining sustainability objectives and measuring outcomes. These connotations also promote more efficient collaboration with other necessary disciplines in the pursuit of project and societal-level goals. The research employs a novel use of computational corpus linguistics in CCE to identify and examine patterns of collocates associated with the word sustainability in a corpus of 335,020 peer-reviewed civil and construction engineering (CCE) scholarly abstracts published between 1981 and 2021. Demonstrating this research method is a contribution to CCE researchers systematically extracting meaning from large volumes of literature.

1. Introduction

The discourse related to sustainability in academic literature has surged in the last 20 years [1]. In civil engineering, the concept of sustainability has also gained popularity in the literature as it addresses today’s needs and impacts on future generations [2]. Despite a significant amount of literature speaking to what sustainability is, how it is approached, and why it should matter, there continues to be much confusion about what constitutes sustainability in the discipline [3,4,5,6]. This confusion leads to inconsistencies in developing standards, criteria, and factors in engineering design, construction projects, and rating systems [7], thus impeding societal progress in a meaningful way. For instance, both Envision (USA) and BREEAM Infrastructure (UK—formerly known as CEEQUAL) are two popular sustainability rating systems for civil engineering and infrastructure projects [8,9]. Both rating systems promote sustainability, yet neither system is compatible with each other due to different focus areas and indicators. The initial step toward addressing this challenge is to gain a comprehensive understanding of how sustainability is portrayed in current publications within the field. However, academics must navigate and make sense of a vast volume of literature on sustainability. This study tackles the issue by using a computational corpus linguistics approach to systematically analyze 335,020 abstracts from peer-reviewed journals and conferences in the civil and construction engineering (CCE) body of knowledge, presenting a comprehensive view of sustainability within the built environment. This approach helps to grasp the extent of sustainability’s presence in the literature, its evolution over time, and its multiple connotations relevant to civil engineers.

The notion of sustainability was initially thrust into the global spotlight by the United Nations in 1987. The World Commission on Environment and Development (WCED) published the groundbreaking Brundtland report, which introduced the concept of sustainable development or sustainability to the world. The report defined sustainable development as ‘development that meets the needs of the present, without compromising the ability of future generations to meet their own needs’ [10]. Later, in 1992, the UN Conference on Environment and Development (the Rio Summit) advocated the same idea [11]. In 2000, the UN’s Development Programme (UNDP) issued eight Millennium Development Goals (MDGs), which were replaced by the 17 Sustainable Development Goals (SDGs) at the UN Summit in 2015 [12,13,14,15]. This series of events paved the way for new scientific and political discussions worldwide as humans try to balance our needs and what nature can supply.

The first abstract containing the word ‘sustainability’ within the Scopus database for the CCE disciplines was in 1991, which discussed the concept of sustainability and its relevance in assessing technology choice issues in irrigation investment appraisals [16]. Since then, many researchers have joined the work from almost every subdiscipline of CCE, in shaping and exploring ways and measures that address sustainability challenges, including motivations and solutions [17,18]. The number of published articles in CCE that refer to sustainability has since grown exponentially. However, disagreements about definitions, the difficulty in understanding the implications of sustainability, and its occasional hijacking for commercial purposes, lead some to question the term [19]. This is particularly true in CCE; with some pointing out that the concept is still contested and evolving [5], and others arguing that it is too complicated to define [20]. Recently, the American Society of Civil Engineers (ASCE) has attempted to define sustainability and sustainable development, in which sustainability is viewed as economic, environmental, and social conditions to maintain society’s quality of life, and sustainable development is the process of achieving those conditions [21]. These definitions clearly follow the prevalent conception of a ‘triple bottom line’ related to sustainability [1], but provide less relevance and interpretation for civil engineers to discern sustainable practices in the field. On the other hand, there have been many attempts to measure the ambiguous and contested term in the built environment through evaluation programs without an explicit definition, such as LEED (US-based), BREEAM (launched in the UK), PASSIVHAUS (German scheme), and World Green Building Council (worldwide) [22]. Despite many overlapping criteria among these programs, there is no universal agreement on what sustainability criteria and measurement schemes should constitute due to the fuzzy definitions of sustainability as a concept.

To understand how civil engineering disciplines employ and interpret sustainability concepts, it is essential to undertake a systematic analysis of the word sustainability used in the CCE scientific literature. The present study utilizes a widely accepted linguistic approach for ascertaining word meaning in semantics and lexicography known as corpus linguistics (CL) [23], which leverages large samples of language (i.e., a corpus) to uncover patterns of meaning and language use [24]. To study the meaning of sustainability, a 62-million-word corpus of peer-reviewed publication abstracts spanning from 1981 to 2021 was constructed. These abstracts were collected from the ‘Civil Engineering’ and ‘Construction and Building Technology’ disciplines as found within the Scopus database, the largest and most comprehensive abstract and citation database of peer-reviewed literature. The corpus is designed to represent contemporary CCE research in the English language. The results of the application of the CL methods of frequency, collocation, and concordance analyses of sustainability reveal the following: (1) the focus of the term has changed and expanded over time; (2) the environmental aspect of sustainability has been the main driving force for sustainability until 2015, with social aspects of sustainability becoming more important in today’s research; and (3) six connotations of sustainability have emerged in the CCE discipline: ‘assessment, ‘progress’, ‘permanence’, ‘abstractness’, ‘multidimensional’, and ‘intersectional’. These six connotations were derived from major collocates of sustainability in CCE literature, and they demonstrate the complexity of sustainability in CCE literature. Current approaches to understanding and promoting sustainability are often through fractured and incomplete lenses. Failing to consider the connotation of sustainability will continue to lead to contested explanations of sustainability as authors seldom conceptualize sustainability in the same ways, which leads to warm conflict among scholars attempting to describe and critique approaches to action. This naturally leads to inadequate solutions to practical problems. These findings promote a holistic understanding and interpretation that systematically addresses sustainability in research, teaching, and practice.

2. Methodology—Corpus Linguistics

Corpus linguistics (CL) is a research method that leverages a large sample of language, known as a corpus (corpora as plural), to understand patterns of meaning in a given context [25]. It is particularly useful for uncovering different uses of words in specialized domains [26]. This is because word meaning is primarily developed as words are repeatedly used by speakers and writers, so by looking at many examples of how the words are used, patterns of use become apparent [24]. CL often considers hundreds or thousands of different texts and with the aid of specialized statistics can reveal meaningful recurrent patterns across many texts and authors as well as different uses across domains. In this research, the authors compiled CCE article abstracts available in Scopus to investigate how the word sustainability is communicated and what meaning this word has carried in the past 40 years within this research field.

In the context of the empirical nature of this study, it was expected that the following research questions could be answered:

• How prevalent is the term sustainability used in CCE research?
• What is the focus of sustainability in CCE research? How does it change over time?
• What meanings of sustainability are pertinent to civil engineers?

The above-listed research questions are answered through word frequency, collocation, and concordance analyses associated with the word sustainability. These answers provide guidance for sustainability education and research and in the design and construction of civil engineering projects in the future.

2.1. Data Collected: Civil and Construction Engineering Abstracts in Scopus (CCEAS Corpus)

To construct this corpus, sampling was specifically targeted toward established and reputable journal articles with high content accessibility in civil and construction engineering. To select candidate publications, the authors used Journal Citation Reports (JCR) 2020 to filter published journals (in English) with an impact factor greater than one in the civil engineering and/or construction and building technology categories [27]. Combined, 140 distinct journals were found within these categories, which were used to create “the corpus”, used in this research (journal titles will be provided upon request). By purposefully sampling these journals, researchers were then able to gather abstracts from the Scopus database in 2021. The ISSN number of each journal identified from JCR was then queried, resulting in 350,968 articles from 1981 to 2021. However, only 96% (335,020 out of 350,968) of the abstracts were available in Scopus at the time of sampling, providing adequate representation of the CCE body of knowledge. The reasons to include abstracts only are twofold. First, an abstract represents the essence of a research paper. If sustainability appears in an abstract, it is more likely to be a central theme in the paper. Second, collecting full texts requires greater computational power and significantly more time to build a corpus. While using abstracts may limit some detail and nuance found in full texts, the Civil and Construction Engineering Abstracts in Scopus (CCEAS) corpus allowed us to sample a full representation of the literature in the CCE disciplines. In the final step of constructing the CCEAS corpus, the authors used R programming [28,29] to clean and organize the data. All words from those collected abstracts were compiled and stored to create the CCEAS corpus.

The total number of texts (i.e., abstracts) included in the corpus is 335,020, and the total number of words (excluding punctuation) is 61,847,272. The mean number of words per text is 169 (SD = 31.3). It is observed that the mean number of words per text is increasing each year, indicating longer abstracts are a trend in peer-reviewed articles for the disciplines. The data were annotated using the default tools for part-of-speech tagging in Sketch Engine [30].

2.2. Word Frequency

Sinclair notes that those who formally study a document will need to consider how often words occur because the frequency of occurrence indicates a word’s importance and usefulness [31]. It is often expressed in a frequency list that records the number of times each word occurs in the text, with higher frequencies indicating more importance and usefulness within a given domain [32]. However, those arithmetic frequencies simply count the number of occurrences; they do not indicate the prevalence of a word in terms of a proportion of the total number of words within the text. Therefore, frequencies are typically normalized to a rate of occurrence, such as per million words.

2.3. Collocation and Collocation Analysis

The term collocation refers to the fact that certain words tend to co-occur more frequently in natural language use than syntax, semantics, and random chance would dictate [33]. If two words collocate with each other, then they co-occur in some way, usually more often than would be expected if all of the words in a corpus were presented in a random order (i.e., co-locate or located near to each other) [34]. For instance, strong wind or heavy rain. The words strong and heavy are collocates for wind and rain, wind and rain are nodes (central word in a collocation), and both heavy rain and strong wind are collocations when a collocate and node are used together. In this example, the two adjectives cannot easily be reversed without sounding unnatural. However, this co-occurrence indicates more than just the formulaicity of expression. It has also been frequently used to gain insight into the denotation and connotation of words, therefore, by understanding connotations through collocates, underlying conceptualizations and beliefs about a target word can be uncovered [34]. In this research, the authors examined the modifiers, which are collocates that are either pre-modifying nouns or adjectives that provide the node with more meaning. Top modifiers (based on collocation statistics logDice) of the noun sustainability are collocates to the word, and they add a description to the noun and indicate certain meanings or aspects of sustainability.

Further, Firth has pointed out “ye shall know a word by the company it keeps” (Firth, 1957a), and he illustrated as such: one of the meanings of night is its collocability with dark (Firth, 1957b). Xiao also argued that collocation plays an important role in disambiguating a polysemous word [35]. Indeed, the meaning of words can be revealed through collocation because of the mutual expectancy between two words [35]. Hence, collocation analysis could help to answer research question three: what is the context-specific meaning of sustainability for civil engineers? Although this analysis is rare in the CCE discipline, other research has shown the contribution of collocation analysis in the clarification of word meaning in different disciplines. Researchers investigated the word cohesiveness in the food industry through collocation analysis in food texture testing literature [36]. Similarly, researchers in the public administration field evaluated the meaning of red tape in policymaking, media, and academia [37]. Researchers in social science explored the meaning of risk using corpus linguistics [38], while law practitioners use corpus linguistics to understand words and other expressions in their ordinary state [39]. In political science, Chilton used linguistics to uncover the political meaning of the word people in populist discourse [40]. However, corpus-based research in CCE is somewhat rare and primarily pedagogically focused on helping students become better writers or identify CCE-specific vocabulary to learn [41,42]. To our knowledge, this research reports the first significant use of CL in semantic analysis of CCE-focused research.

A collocation analysis quantitatively measures which pairs of words have a substantial chance of occurring near one another taking into account the observed and expected frequencies of each word in the dataset using logDice. logDice is a statistical measure bound between 0 and 14 that is commonly used for identifying collocations and expresses the typicality of the collocation [43]. A higher score indicates a greater strength of association between the target word and the collocate, which is due to some semantic relationship. For example, two words can be highly associated because they are part of a fixed expression, or it might simply be that two concepts are tightly associated with one another. The formula for logDice is as follows:

$$logDice=14+{\log }_2\left(2{\displaystyle \frac{f_{xy}}{f_x+f_y}}\right)$$

(1)

where $f_x$ = total number of occurrences of the target word, $f_y$ = total number of occurrences of co-occurring words, and $f_{xy}$ = total number of co-occurrences of the queried item.

2.4. Concordance Lines

Concordance lines are occurrences of a word form in its own textual environment and are the primary means of qualitatively interpreting corpus linguistic results (see [31]. In the present study, concordance lines are generated around the target word sustainability. Concordance lines collect the entire natural context where the queried term appears and can be manually coded to obtain more detailed information about the word in terms of its implication and application based on its situational use. For example, the word promote collocates with sustainability. However, the exact nature of the relationship between the two words is missing. Thus, concordance lines show the whole context to reveal the reasons why the two words collocate, such as the study stresses the importance of this factor F1 as a means to promote sustainability and energy efficiency in the hotel sector. By looking at the collocates alone, it could perhaps have been the case that these two words are associated because sustainability helps promote growth or that something promotes interest in sustainability; theoretically, any number of possible relationships might exist between these two words. However, an analysis of the concordance reveals that, instead, the sentence is concerned with variables that promote sustainability. The concordance lines provide a better understanding of how a word is used in reality. These concordance lines can then be coded using formal qualitative methods to help understand collocations.

2.5. Content Analysis of Collocates

Content analysis is one of the most widely applied qualitative methods to help determine the presence of certain concepts, themes, or patterns of meaning within a dataset [44], and it is particularly useful for the analysis of collocates and other types of corpus data [45,46]. Content analysis is a summarizing analysis of messages that follows the standards of the scientific method, including attention to objectivity-intersubjectivity, a priori design, reliability, validity, generalizability, replicability, and hypothesis testing based on theory [47]. Validity is often established by accurately reporting the process of content analysis, from data collection to emerging themes [48]. In the current study, the authors also grouped collocates according to their content meaning in the discourse rather than their grammatical and syntactic functions. The steps involved in conducting content analysis are reflected in

Table 1. Steps in content analysis.

Steps	Explanation	Description of the Step in This Research
Choosing content	Determine what content will be examined.	CCEAS, civil and construction engineering abstracts in Scopus corpus.
Defining variables	What variables will be used in the study, and how do you define them conceptually?	Connotations of the word ‘sustainability’.
Defining the unit of data for analysis	What unit of data collection will you use?	Collocates of ‘sustainability’.
Creating a codebook	Human/computer coding, in which a codebook and coding form are created.	Codebook was developed through an iterative process that assigns the codes or categories of the collocates based on their intended meaning in the concordance lines.
Data sampling	Sampling	The frequency-based approach draws on quantitative evidence of word co-occurrence in corpora from which collocations are extracted using frequency cutoff scores and collocational association measures [49,50]. High-frequency and high logDice collocates (40 for frequency and 3 for logDice) were selected because they capture a reasonable number of meaningful collocates of sustainability.
Training human coders	Training and pilot reliability	Two human coders were taught the meaning of each category and how those categories emerged to form the codebook. Coders independently coded ~20 collocates by looking at concordance lines to inform them on how the words were used in context. After each round of coding, their results were compared. Discrepancies were reviewed as a group. This process iteratively continued until an agreement was reached between the two coders.
Human coding	Coding; uses at least two coders to establish intercoder reliability.	After an agreement was reached in training, the coders independently continued working on the rest of the collocates based on the training from step 6.
Triangulation	Final reliability; calculate a reliability figure for each variable.	Analyze overlaps of two coders, calculate Cohen’s Kappa coefficient to establish intercoder reliability, and resolve any discrepancies.
Report the results	Tabulation and reporting	Report the results in Table 4 and Table 5

.

3. Research Findings

3.1. Frequency of the Word Sustainability in CCE

The total number of abstracts in the corpus that used sustainability was 6063, which is about 1.81% of the total abstracts. Even with the number of abstracts increasing each year from 1981 to 2021 (Figure 1), the proportion of abstracts mentioning sustainability has also grown rapidly during that same time. Although both the numbers of abstracts, and occurrences of the word sustainability increase over time, the rate of increase is incompatible. As Figure 2 shows, the word sustainability first occurred in 1991 onward and the proportion of use increased sharply upward in 1998, and increased sporadically, with the highest percentage in 2021. Figure 2 shows the growing trend for abstracts that mention sustainability, which shows increasing research efforts on sustainability in CCE. The interpretation is that more sustainability-related research continues to be published, which causes the total number of occurrences of the word sustainability to increase faster than the number of articles published.

3.2. Collocates of Sustainability

3.2.1. Evolving Collocates

As interest in sustainability grows, the number of abstracts focused on sustainability increases. However, the collocates of sustainability are not always consistent, rather representing changes in different time frames.

Table 2. Top modifiers change over time.

Year	No. of Abstracts Included Sustainability	Top Modifier	Second Top Modifier	Third Top Modifier
1991–2000	107	Economic	Urban	Ecological
2001–2005	231	Ecological	Long-term	Urban
2006–2010	581	Long-term	Ecological	Institutional
2011–2015	1575	Environmental	Long-term	Social
2016–2021	3569	Social	Environmental	Financial

shows the top three modifiers of sustainability in different time periods. Research on sustainability was initially associated more with economic sustainability, then the topic expanded to ecological sustainability, long-term sustainability, environmental sustainability, and social sustainability. With the increasing number of abstracts, the expansion trend indicates sustainability is still an evolving concept in CCE and has no definitive meaning as some have indicated [5]. The three pillars of sustainability (social, economic, and environmental) have become ubiquitous; however, this concept is a gradual emergence from early critiques on balancing economic development with environmental and social problems [1]. Table 2 shows that CCE research has focused on ecological and environmental aspects of sustainability since early on (from 1991 to 2021) when the word first appeared in 1991 in CCE literature. It also signals that social sustainability has started to become a stronger focus only in the last ten years. The shifting emphasis in this research demonstrates that the concept of sustainability in CCE is indefinite but is converging toward the popular definition of the three pillars “framework” of sustainability.

3.2.2. Three Pillars of Sustainability

Defining the concept of sustainability is an ongoing process and scientists and researchers continue to endeavor to capture the complex and multifaceted concept [51,52]. Among them, one particular prevalent narrative is the framing of three interconnected “pillars”: environmental, economic, and social [1,53,54]. In the U.S., the ASCE board of directors also adopted the three pillars definition of sustainability in policy statement no. 418 to welcome a vision of sustainable development [21]. The authors further investigated this broad definition in the corpus and found many researchers embraced the concept of multi-dimensional sustainability. But the term ‘three pillars’ does not always occur; instead, ‘triple bottom line’, ‘three aspects’, and ‘three dimensions’ also often appear in the CCEAS corpus. The following concordance lines are examples of this, using keywords in contexts (KWIC), which have been italicized:

This review identifies the social, economic, and environmental pillars of sustainability that can be achieved using automation and control systems.

To meet increasing challenges of triple bottom line sustainability covering social, environmental, and economic aspects.

How water shortage affects economic, social, and environmental aspects, and what the relationship between system sustainability and water shortage is.

Sustainability is quantified in terms of its social, environmental, and economic metrics.

And a continuous assessment of the environmental, social, and economic dimensions of sustainability adapted to the specificities of such projects.

But broader social, economic, and environmental concerns now figure into the idea of sustainability.

The authors further analyzed each of the concordance lines (6236 in total) to identify the frequencies of each “pillar” and found that the social pillar is least often addressed among the three. Since the CCEA Corpus consists of article abstracts with limited word counts (mostly less than 200 words), the chances of fairly frequent collocates appearing in a particular context, or being produced by a small number of authors, are very small. Thus, by counting the frequencies of each pillar or word environmental, economic, and social within those 6236 concordance lines, one can observe the degree to which the three pillars are weighted in terms of frequency (see

Table 3. Frequency of each pillar of sustainability.

Three Pillars	Frequency of Each Pillar
Environmental	1219
Economic	504
Social	501

). Both economic and social components of sustainability are very similar in terms of frequency; however, some have argued that even if economic sustainability is mentioned less than environmental sustainability, it does not mean economic sustainability is less valued [55,56]. Rather, it is generally found to be the opposite, in that economic sustainability has a paramount impact on civil engineering projects and becomes a prerequisite in design and construction [3].

3.2.3. Social Sustainability

Social sustainability is a relatively new concept for the CCE disciplines, the origin perhaps lies with the three pillars of conceptualization. The authors analyzed 535 concordance lines containing both social and sustainability and found that 279 (56%) concordance lines contributed to the multi-dimensional (mostly three) construct of sustainability. The implied assumption is that the increased discussion on social sustainability is mainly due to the increased discussion of the multi-dimensional concept of sustainability. On the other hand, researchers also criticized the lack of social sustainability studies in the CCE disciplines and the need to strengthen this work [56,57,58]. The following concordance lines (KWIC in italics) revealed this concern:

While the social pillar of sustainability is not addressed to a significant degree.

Honors the commitment of the SDG goals to take equal account of social and sustainability concerns.

Social dimensions, as one of the three pillars of sustainability, are commonly not fully considered in the planning, bidding, and design processes of construction projects.

…are expected to achieve sustainability economically, socially, and environmentally. Despite this, their social sustainability level is relatively low and is yet to be improved.

Therefore, social issues were less included in the main categories of sustainability assessment schemes.

3.3. Corpus-Based Definition of Sustainability in CCE

Sketch Engine was used to collect and analyze the collocates of sustainability, in which it selected words, three left and three right, around the KWIC. The total number of collocates found in the corpus for sustainability is 1166. The highest frequency (the) is 2500, while 966 collocates have frequencies below 20. Figure 3 is a distribution of the collocates (with frequency higher than 10) based on their frequency in the corpus, the long tail indicates many collocates with lower frequency.

The higher logDice score and frequency indicate a more substantial relation between the KWIC (in this case sustainability) and the collocates. Therefore, this brings more information to the KWIC that can suggest definitions or other useful perspectives of the word sustainability in civil and construction literature. This research used a combination of frequency cutoff (40) AND logDice score (3) as the threshold. Any collocates with a frequency lower than 40 AND logDice score lower than 3 were excluded from further collocation analysis. This resulted in exactly 100 collocates. Although the lower frequency and logDice score hypothetically could reveal additional information about sustainability, it is difficult to conclude that there are meaningful patterns.

Content Analysis of the Collocates

After compiling these 100 collocates, a codebook was developed for the coders to use during this content analysis process. The codebook structure was developed initially by using an inductive approach and then iterated until reaching a consensus of categories for the shared collocates and concordance lines from the corpus [59,60]. Through three iterative rounds, the team ultimately created a codebook that contained nine themes/categories with definitions and example collocates as shown in

Table 4. Codebook with definitions.

Themes/Categories	Theme Definition	Example Collocates with Raw Frequency
Evaluativeness	Words that describe sustainability can be measured and evaluated.	assessment (284), performance (213),
Aspiration	Words that imply sustainability as a positive or desired goal.	improve (268), achieve (155)
Natural Environment	Words that indicate things that naturally exist in the environment.	environmental (407), water (166)
Built Environment	Words that involve the human-built environment.	building (246), urban (237)
Abstractness	Nouns that often describe means that are either ideas or theories and, therefore, are abstract or intangible.	project (110), issues (104)
Multi-dimensional	Words that indicate multiple aspects or dimensions.	system (125), aspect (112),
Resilience	Words that indicate long-lasting.	long-term (146), resilience (138)
Economics	Words relate to economics and cost.	economic (154), efficiency (83),
Others	Words that have conjunctional functions that either do not provide information or provide little details about sustainability.	the (2500), of (2426)

. The nine categories demonstrate a diverse usage of the term sustainability in CCE literature. This varied representation could potentially lead to confusion and even contested interpretations regarding the concept of sustainability.

Two independent human coders were trained and carried out the coding following the codebook. The total overlapping rate was found to be about 96%, and Cohen’s kappa is 0.955 (high reliability), both indicating the degree of the common agreement is high enough to be considered reliable coding. As far as dissimilarity, it lies in these four collocates: system, ensure, industry, and index. These collocates were reexamined at the concordance level and were recategorized after reaching a consensus between the coders. Because the ‘others’ category contains collocates revealing less meaning for sustainability in CCE (i.e., be, have, its, their, etc.), it was removed from further analysis.

After the collocates were grouped according to the categories and themes, the total number of collocates and frequencies were added up in each category. The number of collocates in each category shows different ways researchers approach similar aspects or characteristics of sustainability. The total number of frequencies in each category suggests how often researchers address that aspect or characteristic, which also signals how important the aspect is to CCE researchers. In a way, higher frequency purports a higher degree of importance to CCE researchers.

Table 5. Summary of the categories.

Themes/Categories	Collocates	Total Number of Collocates	Total Number of Collocation Frequencies
Evaluativeness	assessment (284), performance (213), assess (159), indicator (142), index (124), criteria (118), evaluate (82), tool (79), measure (68), evaluation (65), analysis (48)	11	1382
Built Environment	building (246), urban (237), social (208), construction (183), concrete (85), city (82), infrastructure (80), transportation (64), built (44), industry (40)	10	1269
Aspiration	improve (268), achieve (155), increase (113), enhance (101), goal (79), ensure (59), promote (58), important (42)	8	875
Natural Environment	environmental (407), water (166), energy (119), resource (85), groundwater (70)	5	847
Abstractness	project (110), issue (104), design (80), concept (79), approach (74), impact (69), model (59), principles (50), objective (49), management (48), concerns (46)	11	768
Multi-dimensional	system (125), aspect (112), levels (78), framework (60), dimensions (58), three (50), different (48)	7	531
Resilience	long-term (146), resilience (138), durability (46)	3	330
Economics	economic (154), efficiency (83), cost (65)	3	302
Others	of, the, and, to, in, for, a, is, on, this, are, as, that, with, use, their, by, can, its, towards, an, has, from, development, be, through, it, into, however, such, paper, which, based, at, have, while, these, contribute, between, also, about, overall	42	13,060

is a summary of the categories with their respective frequencies and number of collocates. It shows that one of the most important characteristics of sustainability in CCE is ‘Evaluativeness’, which reflects that researchers are keenly interested in how to evaluate and measure sustainability. The category of the second-highest total frequencies is ‘Built environment’, which sets the boundary where civil engineering projects operate. The third category is ‘Aspirational’, indicating sustainability is a concept that researchers welcome and embrace. The categories ‘Natural environment’ and ‘Built environment’ illustrate that civil engineering projects are at the nexus of both environments and the civil engineering discipline plays a vital role in creating a sustainable environment. ‘Abstractness’ is another characteristic that shows the concept of sustainability is intangible. These intangible words describe sustainability’s elusiveness as an ideological concept in the discipline [61]. The ‘Multi-dimensional’ category shows the term has multiple facets or components. Scientists have long before recognized sustainability as a plural, multi-dimensional approach [1,62] The ‘Resilience’ category contains collocates that imply sustainability is long-lasting. Resilience and sustainability are often interchangeable in many literature [2,63]. In civil engineering, the common view on resilience is the ability that structures to withstand deterioration and adapt to hazards, which is a component of achieving sustainability [64]. This category shows the closeness between the two words in the CCE discipline, which supports the aforementioned views on them. The ‘Economics’ category has the least number of collocates and collocation frequencies; apparently, researchers did not investigate much in the economic aspect of sustainability. This lack of research on economic sustainability does not imply the concept is less valued in the discipline, it is rather the contrary since economic conditions are dominating the decision-making process pertaining to civil engineering projects [2,56].

4. Discussion

The purpose of this research was to investigate the sustainability concept within CCE disciplines, specifically; the prevalence of the sustainability concept, how the sustainability concept has changed over time, and what the main themes are associated with the sustainability concept. The novelty of the research method and technological improvements in computational power have allowed researchers to reveal hidden patterns and usage of sustainability in CCE that have yet to be studied. The three primary results are (1) CCE researchers have increased their focus on sustainability, (2) this focus on sustainability is changing, and (3) sustainability can best be understood through six distinctive connotative conceptions.

4.1. Increasing Focus on Sustainability

This result is not overly surprising, it is well aligned with the contemporary paradigm shift to a more sustainable world in today’s society [1]. However, the innovative use of corpus linguistics allows us to demonstrate an exponential increase in the number of CCE academic publications addressing sustainability between 1981 and 2021. In light of this proliferation of scholarly publications, it is evident that there remains a substantial body of work yet to be cultivated within this domain. For instance, how to balance each aspect to achieve holistic sustainability [65], lack of design guidance on sustainable structures [66], and developing sustainability by using artificial intelligence AI systems [67], etc.

4.2. Sustainability Trends

This research has also revealed that the scope of sustainability has been shifting and expanding over the last 40 years within the CCE disciplines. Water sustainability was studied most in the early days when sustainability and environmental sustainability emerged, as such studies often included water and other natural resources in the natural environment. In fact, environmental or ecological have been the most frequent modifiers for sustainability in the early years. Additionally, their saliency ranks the highest among the years indicating discussion dominance across CCE. However, the saliency of social has exceeded the saliency of environmental from 2016, suggesting that the focus is shifting more on humans in today’s society. In fact, the top three modifiers from 2016 to 2021 were social, environmental, and financial. These are closely corresponding to the ‘three pillars of sustainability’, one of the most popular sustainability concepts from the UN. The occurrence of social heavily overlaps with the occurrence of the three pillars concept. An implicit notion is that the three pillars concept currently dominates the discussion of social issues in CCE research.

4.3. Six Connotations of Sustainability

To answer the research question concerning the applicable meanings of sustainability within the field of civil engineering, we further reviewed and refined the eight collocate categories to produce six major connotations that emerged in the literature. The connotations are the representative intention and meaning of those themes for civil engineering, in order to make the content useful for researchers and practitioners. The diversity of these connotations emphasizes that sustainability does not have a singular meaning for CCE researchers. It is common for sustainability to be perceived as ambiguous or subject to debate across various academic works, as evidenced by multiple studies [68,69,70,71,72,73,74]. This inherent ambiguity poses a challenge in advancing the attainment of sustainability. These connotations not only shed light on the reasons behind this ambiguity but also offer multiple perspectives of sustainability that can guide civil engineering researchers in addressing this concept. These connotations represent the distillation of those recurring themes/categories derived from the content analysis of frequently occurring collocates in the corpus. These connotations are described and discussed below.

Assessment—sustainability requires evaluation and measurement of CCE practices; see ‘Evaluativeness’ in Table 5.

Assessment tools can be instruments to improve the efficiency of design and construction processes [75]. These assessment tools provide guidance on meeting required standards and specifications and help the discipline address sustainability issues in systematic ways. Civil engineers should be intentional when selecting and applying assessment tools.

Progress—the general attitude towards sustainability is to improve communities and societies for current and future generations; see ‘Aspiration’ in Table 5.

The concept of sustainability gained prominence in 1987 and has since evolved into a central theme in global development. It encapsulates the idea of addressing present-day needs while also taking into account the consequences for future generations. This connotation reflects a prevailing sentiment and signifies a path toward achieving a sustainable world. In the field of CCE, it is crucial for researchers to further enhance our understanding of sustainability and expand its application throughout the civil engineering process, encompassing design, development, operation, and decommissioning.

Permanence—sustainability should lead to more durable, resilient, and more permanent action; see ‘Resilience’ in Table 5.

Permanence denotes the extended duration of a structure or system in CCE, a critical facet in addressing sustainability. This attribute of sustainability pertains to civil engineering endeavors with enduring societal implications, including infrastructure like energy, water, and transportation systems.

Abstractness—sustainability can be abstracted to differing levels of analysis and is conceptualized in distinct ways in different disciplines; see ‘Abstractness’ in Table 5.

Sustainability, as a concept, lacks a physical presence and is rooted in general principles and thought processes [76]. This presents a particular challenge for applied disciplines like CCE. Progress in this field will be impeded until civil engineers can concretely establish applicable standards and tools for implementing the concept.

Multidimensional—sustainability is multidimensional as it relates to environmental, social, and economic impacts; see ‘Natural environment’, ‘Multidimensional’, and ‘Economics’ in Table 5.

An effective way to understand the concept of sustainability is to break it down into different aspects. Civil engineers acknowledge diverse perspectives of the concept through which sustainability is approached. This recognition prompts engineers to navigate the various components of sustainability without showing a pronounced bias toward any single aspect [21].

Intersectional—any focus on sustainability in the CCE disciplines requires cognizant attention to the intersections between the natural, social, and built environments; see ‘Built environment’ and ‘Natural environment’ in Table 5.

Civil engineers hold a vital position in building societal infrastructure, situated at the crossroads of the natural, social, and built environments. This interconnected role highlights the crucial need to harmonize these domains and navigate their complex relationships. This connotation of sustainability emphasizes the significance of integrating all three domains and addresses their intricate relationships in civil engineering.

5. Contributions

The term sustainability has been used in numerous disciplines and a variety of contexts, and the meaning of the term is strongly dependent on the context in which it is applied [77]. A lack of consistent definitions for sustainability in literature also poses a great challenge to the research community [52]. This challenge is real in the CCE discipline; a lack of definitional clarity leads to fractured approaches and misaligned actions among stakeholders. In order to understand how the term was used in CCE, we examined past research related to sustainability in the literature and found trends and patterns of the term that contribute to the body of knowledge in three different ways.

5.1. Theoretical Contribution

The primary theoretical contribution of this research is a comprehensive description of sustainability beyond the three-pillar conceptualization for CCE researchers. We recognize that the three-pillars concept is a useful heuristic, but the ‘pillars’ largely represent the ‘domains’ of sustainability. There is more nuance than what the three-pillar concept provides concerning CCE disciplines when addressing sustainability in a comprehensive way. CCE sustainability research will continue to address fundamental principles of sustainability beyond the domain of action (three pillars) and also include the purposes, approaches, and specific thinking captured in the six connotations. Using research language in the CCE discipline, we provide researchers, educators, and practitioners with nuanced backgrounds from which to base future statements, scholarly communications, and theoretically rigorous operationalization of sustainability.

5.2. Methodological Contribution

Corpus linguistics is often used in language acquisition and language interpretation. It essentially explores and examines the patterns of language in discourse, providing various features of word usage that could easily be overlooked by other data processing methods [31]. However, this approach has rarely been used in civil and construction engineering. Although some literature research in CCE has been done via text-mining [78,79,80], each studied a small set of texts. Despite the infrequent use of CL outside of the linguistics and law disciplines, it is well suited to mega-text data analysis regardless of the nature of the disciplines. We recommend a wider application of the methodology in CCE to provide new and unique insights.

5.3. Practical Contribution

The practical contribution of this research lies in the proposed six sustainability connotations derived from the CCE literature. This work supports previous efforts where collocational network analysis was used to demonstrate that an alternative paradigm of three interconnected environments (natural, built, and social) is more appropriate for the CCE sustainability context [81]. This research generated an additional nuanced understanding of sustainability for the CCE context, and the six connotations allow civil engineers and construction managers to better address sustainability following a more principle-centered approach. Application of these connotations can broaden the way project assessments are conducted. Additionally, the connotations can inform business and governance decision-making for more permanent and resilient outcomes. They also promote more efficient collaboration with other disciplines on sustainability issues. The concept of sustainability provides a holistic solution to many of the crises humans face today, such as energy shortage, pollution, resource scarcity, climate change, social justice, etc. Civil engineers work to address these crises and are central to the nexus of the natural, social, and built environments. Civil engineers have the potential to make an immense impact in pursuit of sustainable communities and society. This impact is achieved by advocating for and leading in, sustainable design and practice. Civil engineers must be adequately trained and equipped with the intellectual and technical knowledge to be leaders in solving real-world problems in the built environment. We call for the application of these nuanced connotations of sustainability to promote more comprehensive thinking that leads to holistic solutions and guidance in solving challenges humans face today, and into the future.

6. Limitations and Future Work

This research used article abstracts from the Scopus database to construct the CCEAS corpus. Although the corpus is large enough to yield meaningful patterns and results, limitations associated with its sampling should be considered. There is other published scholarly work outside the Scopus database. Although abstracts are brief summaries of published papers, they lack the detail contained in full-text papers. Additionally, while the descriptive statistics of our collocation analysis were generated by our algorithm, the underlying themes (content analysis) to which the collocations belong were interpreted by the researchers themselves. This interpretation, as well as their implications, ensures that research findings are placed within their logical context, but includes a certain degree of subjectivity. Future research may extend and strengthen this work by (1) building corpora that include full-text articles containing sustainability and conducting additional corpus analysis, and (2) building corpora that include research from disciplines related to CCE and conducting cross-discipline comparisons on sustainability.