Abstract

As a specialty area, criminology is a constellation of many different interrelated subfields that study the various institutions constituting and person(s) involved in the criminal justice system. While the researchers that contribute to these subfields are independent, the contributed research is inherently interdependent – the development of criminal behavior is influenced by contacts with law enforcement, the courts, corrections, and the community. Consequently, cross-pollination between subfields is necessary to ensure innovative research that can be translated into evidence-based interventions that positively affect offenders, victims, and criminal justice stakeholders. In this paper, we synthesize topic modeling and main path analysis (MPA) to analyze criminology and criminal justice abstracts and metadata from peer-reviewed research published between 2001 and 2020 (N=31,257), investigating the degree of overlap between life-course criminology and policing research. This weight activated main path analysis (WAMPA) uses measures of relevance measures to ‘steer’ MPA towards focal topics, in this case life-course and policing research. In so doing, we map the most influential policing and life-course literature from the last two decades, then examine citations that directly connect the two concurrently developing bodies of work. We find very little cross-citation between the main paths of policing and life-course criminology, almost exclusively received by flagship publications. We further find that cross-citations typically represent methodological proliferation, and fundamental disagreement, with little evidence of prominent theoretical and empirical integration.

Introduction

As a specialty area, criminology is a constellation of many different interrelated subfields. Generally speaking, each area focuses on the institutions which constitute the justice system, the individuals that comprise those institutions (e.g., police), or the individuals that encounter the system (e.g., offenders). While the researchers that contribute to these subfields are independent, the contributed research is inherently interdependent. The development of criminal behavior is a downstream confluence of individual-level factors, as well as external experiences such as contact with law enforcement, the courts, corrections, and the community. This is important for a variety of reasons, but it is particularly salient when attempting to translate innovative scientific research into evidence-based, ethical policies (Leahey & Moody, 2014). While one might not dispute the importance of dialogue across disciplines, the extent to which meaningful communication actually occurs is an entirely different question. Criminal justice researchers may exist in the same professional space, but social scientific research is still frequently conducted in relative isolation (Leahey, 2016). Scholars testing criminological theories tend to be represented by different semantic and linguistic features than scholars who study the criminal justice and legal systems (courts, corrections, policing, etc.). Collaboration between these subfields tends to be infrequent and varied (Smith et al., 2023a). Life-course and policing research, in particular, have matured into dominant subfields characterized by distinct linguistic paradigms over the last several decades (Laub & Sampson, 2020; Weatheritt, 2023). Some conceptual collision between the two areas might seem unavoidable; for instance, the age of offending onset is a topic central to developmental research, with the obvious overlap being that this point in the life-course often marks the first official encounter with law enforcement (Moffitt, 1993). Nevertheless, the conceptual brushes between these fields is yet to have a discernable impact (see Novak et al., 2023). Using English-language, peer-reviewed publication data collected from dimensions.ai, we construct a citation network and perform main path analysis (MPA) with the goal of identifying the ‘core’ of the field – the research to which many scholars trace their criminological lineage. Having identified the ‘core’ of the field, we then build on the work of Chen et al. (2022), introducing a novel approach to guiding MPA along specific topical pathways using the topic embedding model, top2vec (Angelov, 2020). In so doing, we identify the main paths of both life-course and policing without omitting criminology’s broader context. This allows us to examine the concurrent development of policing and life-course criminology, including points of divergence and convergence.

Criminological collaboration and the science of science

The science of science (SciSci) is a transdisciplinary field, adjacent to Informetrics and Scientometrics, that analyzes data produced by academics and scholars in an effort to understand the mechanisms underpinning the ‘doing of science’, and develop tools and policies that can help improve and accelerate this process (Fortunato et al., 2018). Collaboration between (sub)fields within the scientific domain is vital in fostering innovation, addressing complex problems, and maximizing resources and expertise (Garfield, 1979). Interdisciplinary discussions help bridge the silos between different scientific disciplines, encouraging the cross-pollination of theories and methodologies (Davies & Ruda, 2023; Merton, 1973). Criminology began its life as an outgrowth of sociology, yet relatively few decades passed before adjacent fields like psychology, law, and biology permeated the discipline (Cressey, 1979).

The most frequently cited reason for collaboration in SciSci scholarship is specialization (Leahey, 2016). Specialization is viewed as a necessary reaction to a rapidly growing body of scientific knowledge (Evans, 2008; Landhuis, 2016). Scientific collaboration is thus viewed as a necessary extension of this individual specialization, where certain ‘specialty areas’ – including criminology, criminal justice, legal studies, and the specializations therein – can require the synergistic interaction of scholars from a variety of different fields and subfields (Leahey, 2016; Smith et al., 2023a). However, this integration and collaboration is easily obstructed by administrative, geographic, and even linguistic boundaries between different disciplines (Biancani & McFarland, 2014; Brint, 2005; Smith et al., 2021). Catalyzing and sustaining communication can help overcome these boundaries, break down (sub)disciplinary silos, and catalyze innovative solutions to socioeconomic issues that are beyond the scope of any individual subfield (Smith, et al., 2023b).

Inter- and intra-disciplinary collaboration and communication is especially important in a specialty area like criminology. Crime and the criminal justice system beget issues of human development, victimization, deviant behavior, community development, economic structures, historic systems of power, legal theory, etc. (Walsh & Ellis, 2006). To develop effective public policies to combat crime, researchers within the field of criminology must communicate with each other and others. However, even as a discipline comprised of myriad interdependent subfields (Garland, 2011), criminology has historically been characterized by the dominance of certain perspectives and ‘paradigms’: sociological criminology throughout the 1900s (Jeffery, 1978); developmental and life-course criminology with the turn of the century (Laub, 2004). The extent that these dominant paradigms and the many surrounding subfields are communicating with one another is not well understood, and neither are the boundaries that might exist within criminology.

The intersection of life-course and policing

Though empirical studies are limited, there are theoretical and conceptual reasons to assume that criminology’s subfields are actively communicating. Among the divisions of the American Society of Criminology (ASC) and the Academy of Criminal Justice Sciences (ACJS), corrections and sentencing overlap with policing by virtue of their adjacency in the criminal justice system and similarities in their practitioners (Liebling, 2000), victimology overlaps with corrections, sentencing, and policing by virtue of the victim-offender overlap (Berg & Schreck, 2022), whereas rural criminology is broadly relevant to most other divisions (Donnermeyer & DeKeseredy, 2013). Developmental and life-course criminology is similar to rural criminology in that it is broadly relevant to most other subfields. The life-course perspective is generally considered to be a theoretical perspective and approach to research, examining how stability and change influence deviant behaviors across the life-course, with consideration of the role of developmental stage and age in its examinations of offending (Elder, 1985; Sampson & Laub, 1990). The life-course perspective is thus an inherently flexible theoretical framework, making it compatible with a host of other subject areas in criminology (Krohn & Eassey, 2014). For example, the life-course perspective can complement general strain theory (Agnew & Brezina, 2019), or be combined with biosocial perspectives (Moffitt, 1993), or some combination of all three. These syntheses facilitate an understanding of how individuals respond to their environments and how these responses may differ according to age and developmental stage as well as biosocial settings and interactions (Connolly & Beaver, 2015). The life-course perspective is also compatible with corrections scholarship. For example, the risk-need-responsivity principle suggests correctional interventions are most effective when they are matched with an individual’s presenting risks, needs, and levels of responsivity (Andrews et al., 2011). The life-course perspective suggests these risks, needs, and responsivity may be conditioned upon an individual’s stage in their life-course, dependent upon age and development as well as existing bonds and systems of informal control (Brogan et al., 2015).

The relevance of the life-course perspective to policing scholarship, policy, and practice has become increasingly perceptible. As one of the most forward-facing aspects of the criminal justice system (Reiss, 1971), police regularly come into contact with individuals at all stages of the life-course (Tapp & Davis, 2022). Annually, millions of individuals experience police- and citizen-initiated police contact, from childhood to later adulthood (Puzzanchera et al., 2023; Tapp & Davis, 2022). Given that these contacts will be with citizens suspected of committing a crime, victims of a crime, or individuals looking for help, knowledge of how age and life-course stage influence and impact individuals over time seems an important component of police research, practice, and training. Indeed, there are many areas in which policing and life-course research overlap. Procedural justice scholars’ concurrent interest in police contact (civilian focus) and police practice (officer focus) implies active collaboration and communication (Granot & Tyler, 2019; Murphy, 2015). Work on the victim-offender overlap is similarly situated (Beckley et al., 2018; Bucerius et al., 2021), suggesting researchers in this area may occupy a unique, intersecting topical space. Researchers studying school resource officers may also represent a potential overlap between life-course and policing scholarship, with school resource officers existing at the intersection of adolescent development and policing (Theriot, 2016; Theriot & Cuellar, 2016).

Yet, while the substantive relevance of policing work to life-course scholarship is reasonably well understood, the relevance of life-course work to policing scholarship is not immediately apparent. Research on the adverse consequences of police stops is generally framed within a labeling perspective (Wiley et al., 2013), contributed by researchers with a life-course background (see McGlynn-Wright et al., 2022; Turney, 2022; Wiley et al., 2013). Cumulative disadvantage work is similarly framed to emphasize the relevance of policing to the life-course (Kurlychek & Johnson, 2019). Policing scholarship necessarily focuses on police operations and practice. Scholarship at this intersection reflects this focus, tending to borrow theoretical concepts from the life-course perspective to better understand police officers and organizations (Howes & Goodman-Delahunty, 2014; King, 2009), and the factors that influence attitudes towards, and drive engagement with, law enforcement (Bosick et al., 2012; Schuck, 2013). Nevertheless, integration remains limited, and the full extent that scholars from each subfield are actually collaborating in an effort to integrate policing and life-course scholarship is not well understood. Given the broad relevance of both subfields of criminology and criminal justice, it is reasonable to assume that there are myriad untapped opportunities for further integration, and lessons that both fields could learn from the other. Moreover, the efforts of the scholars working in these intersections may not spark broader shifts. Life-course and policing scholars could still occupy separate scholarly silos, necessitating a deeper examination of communication between policing and life-course research. 

The importance of life-course and policing collaboration for policy

Communication between policing and life-course scholars is important for the evolution of the discipline but is arguably more important for the creation of evidence-based, equitable policing policy and practice. At the most basic level, policing practice and training intersect with life-course work through police encounters. Police come into contact with individuals of all ages (Puzzanchera et al., 2023; Tapp & Davis, 2022). The appropriate action or actions taken by a law enforcement officer is dependent on the developmental stage of both offender and victim. For example, in a given shift, an officer may encounter a child accused of stealing a candy bar, a teenager skipping school, a middle-aged adult caught speeding, and an elderly individual who ran a stop sign. Each of these situations reflects the unique intersection of life-course research and policing practice. Understanding life-course tenets would better inform officers of the appropriate response. Appropriate responses for a middle-aged adult, for example, may not be appropriate for a child or teenager, nor may responses for the teenager be appropriate for an elderly adult. While police officers may rely on experience and intuition to navigate these scenarios (Abrams et al., 2020; Huff, 2021), research also suggests police intuition regarding the relationship between age and behavior is heavily influenced by stereotype bias and perceptions of culpability and competence, both of which may be conditioned upon a suspect’s age and race rather than evidence (Abrams et al., 2020; Goff et al., 2014; Perillo et al., 2023).

Given the potential for individual-level biases (implicit and/or explicit) to shape developmentally-informed responses to individuals of various ages, some life-course scholars have begun advocating for the creation of developmentally-informed training and policies for law enforcement (Abrams et al., 2020; Novak & De Francisco Lopes, 2023, 2024). Though officers who are assigned responsibilities as school resource officers receive some training on adolescent development (Counts et al., 2018; Weiler & Cray, 2011), this training is often limited to adolescent development and offered to officers in school settings only, rather than provided as a requirement for all officers (Novak & De Francisco Lopes, 2023). Adolescence is certainly an important stage of human development vis-à-vis deviant behavior (Farrington, 1986). However, it is not the only important developmental stage, nor is training on adolescent development sufficient to provide officers with insight into how to navigate interactions with individuals of all ages, indicating a greater need for life-course and policing research to communicate and collaborate.

Police policy presents another opportunity for possible overlap between life-course and policing work. In recognition of the varying individuals officers interact with on a daily basis, many law enforcement agencies direct officers to consider factors like age, culpability, and competence in deciding whether to arrest an individual or how to address a situation (Abrams et al., 2020; Novak & De Francisco Lopes, 2023). Certainly, directing officers to consider age and its possible impact on culpability and competence is important. However, in the absence of training, the direction may not result in any tangible differences in officer responses to individuals based on their age or developmental stage (Novak & De Francisco Lopes, 2023). Closer considerations of age do little to mitigate individual-level biases that can affect how officers perceive age and its relationship to competence and blameworthiness (Goff et al., 2014; Perillo et al., 2023). It also does little to inform officers how age should be considered so that they might conduct themselves appropriately. Though some may consider age-based responses to be intuitive, research suggests discretion in this area allows for race and gender biases in decision-making (Goff et al., 2014; Perillo et al., 2023), opening the door for developmentally-inappropriate responses to victims and individuals suspected of engaging in crime (Abrams et al., 2020). Absence of clear communication and collaboration among life-course and policing scholars has the potential to negatively influence police policy regarding age and developmentally informed conduct. For all of these reasons, gaining some clarity about the nature of communication between policing and life-course scholars remains a vital task.

Data and Methods

We analyze English-language peer-reviewed publications indexed as “Criminology” under the Australian and New Zealand Standard Research Classification (ANZSRC) Fields of Research (FoR) (Australian Bureau of Statistics, 2020) by dimensions.ai, from which we downloaded the data via the application programming interface (API). The initial sample consisted of 34,551 peer-reviewed publications after dropping 21,240 cases with missing abstracts[1], and 9,744 authored by scholar(s) with fewer than 3 publications and no ORCID profile[2]. Included in the dimensions.ai metadata are the citation lists associated with each publication. Using the igraph R package we construct a citation network where each node is a publication in criminology and criminal justice, and each edge is a directed, acyclic citation where the node projecting the edge is citing the node that receives the edge (Csardi & Nepusz, 2005). After eliminating isolates, this network consisted of 31,257 nodes (publications) and 183,209 edges (citations).

To identify life-course and policing publications within the citation network we queried a top2vec model pre-trained on the same initial sample of peer-reviewed publications (see Smith et al., 2023a). To provide a brief overview: ahead of training, all abstracts were preprocessed (tokenization, lemmatization, and part-of-speech tagging) using the spaCy python library (Honnibal & Montani, 2017). Prior to topic modeling, stop words (high frequency words that contain little or no information, e.g., a, I, me, my, we, because, so) and all words that were not tagged with a Penn Treebank POS that is indicative of topical or substantive importance were dropped from the corpus (Taylor et al., 2003)[3]. Top2vec uses shallow neural networks to ‘learn’ numeric representations of words, documents, and topics by repeatedly attempting to predict adjacent context words for all unique words in a corpus, takes the network weights (embeddings) to represent the target words and documents, reduces the volume of embeddings using UMAP, and identifies coherent clusters of embeddings (i.e., topics) using HDBSCAN (Angelov, 2020; Campello et al., 2013; McInnes et al., 2020). We utilized the model’s ‘semantic search documents by keywords’ functionality to estimate the probabilities that publications in the corpus can be classified as “life-course” or “policing”. In lay terms, the model converted a small number of imputed keywords into embeddings – numeric semantic representations of the words – and returned the cosine similarity of those keywords to all documents in the corpus[4]. This produced two variables (node attributes) that measure the topical and semantic relevance to life-course and policing, respectively – we refer to this as the relevance measure.

Main Path Analysis

Having embedded the content of documents in the corpus and constructed the citation network, the next goal is to identify the “main path” at the core of criminology and criminal justice. Main path analysis (MPA) represents a collection of algorithms that chart the evolution of research based on the tendency of citations to funnel towards one or more ‘main paths’ (Hummon & Dereian, 1989). Intuitively, this method is an instantiation of the proverb “all roads lead to Rome”, identifying the most frequently traversed pathway(s) through a citation network, to which most or all other citation pathways can be traced. For this analysis, we implement the search path link count (SPLC) and key-route search algorithms. SPLC assigns weights to the otherwise unweighted citation network, representing the number of traversals that flow through each edge in the network when iteratively tracing all possible paths from source nodes (publications that receive citations but do not cite others) to sink nodes (publications that cite others but are not themselves cited) (Hummon & Dereian, 1989). A key-route search then identifies the edge with the highest number of traversals (largest SPLC), searches forwards through the citation network until reaching sinks – identifying the paths with the largest summated SPLC weight – then searches backwards through the citation network with the same goal (Liu & Lu, 2012). This ensures that the edge with the single largest traversal count in the citation network is included on the identified main path. For this analysis, which is interested in the general flow of information in criminology and criminal justice, we perform a global search, identifying paths with the largest summated traversal count, rather than a local search, which selects the next incident edge with the largest weight (Hummon & Dereian, 1989)[5].

Weight Activated Main Path Analysis

The main challenge in the empirical assessment of convergence or divergence between two subfields is successfully guiding MPA so that it will prioritize a specified topical path. A coarse solution offered by most research applying MPA is to simply subset a body of literature based on an indicator of topical relevance – e.g., publication in a specific set of journals (Holt et al., 2016). While this is appropriate for larger, clearly defined disciplines, subfields such as ‘policing’ and ‘life-course criminology’ are not exclusively associated with any single collection of journals within a single discipline. A subset of articles in any given criminological journal could cover either one of these subfields, and research in these subfields – life-course in particular – can extend far beyond traditionally defined ‘criminology & penology’ outlets.

Another solution to this problem is the implementation of topic modeling as a means of identifying the subfield(s) of interest (Smith et al., 2023a). Once identified, the larger body of work (e.g., criminological research) would be subset into the relevant subfields (e.g., policing or life-course). The limitation of this approach – in the context of MPA – is divorce from disciplinary context. As illustrated by Figure 1, the main path along a given subfield may rely on papers beyond the boundary of that subfield for its continued development. There are many examples of highly influential articles published within one area of criminology that influence the development of other areas (e.g., Cohen & Felson, 1979; Gottfredson & Hirschi, 1990; Sampson & Laub, 1993).

Chen et al. (2022) propose semantic main path analysis (sMPA). sMPA is an alternative approach to weighting a citation network ahead of identifying the main path. The edge between the $i^{th}$ and $j^{th}$ node in the graph is reweighted by a linear combination of the semantic similarity between the publications those nodes represent, $W_s(E_{i,j})$, and the topological weight (e.g., SPC) calculated during MPA, $W_t(E_{i,j})$, or:

$W(E_{i,j})=a*W_s(E_{i,j})+(1-a)*W_t(E_{i,j})$

where $α$ is a hyperparameter intended to allow the user to prioritize either of the two weights during estimation (Chen et al., 2022). sMPA is thus intended to extract a main path where any two constituent publications are semantically similar, ensuring focus on a single, coherent topic – ostensibly avoiding paradigm shifts. While sMPA can ensure paths in the citation network are relevant to a single topic, it does not allow the user to intentionally guide MPA towards a topic of interest. This requires weighting the network based on similarity of each node to a pre-defined research topic.

The method we introduce, weight activated main path analysis (WAMPA), achieves this through (a) summation of the semantic similarity of each node (publication) to a pre-defined topic across all connected dyads in a citation network, and (b) the implementation of an activation function that ‘switches off’ topically irrelevant edges (see Figure 1). First, we extract document and keyword (topic) embeddings from the top2vec model introduced previously. Using this information, we calculate a topic relevance score for both ‘life-course’ and ‘policing’:

$cos(document,topic)=\frac{\sum_{i=1}^n Document_iTopic_i}{\sqrt{\sum_{i=1}^n(Document_i)^2\sum_{i=1}^n(Topic_i)^2}}$

where $cos(document,topic)$  is the standard parameterization of cosine similarity between a vector of document embeddings, $Document_i$, and a set of topic embeddings, $Topic_i$. Cosine similarity is interpreted as one would a correlation coefficient, varying between -1 and 1. Scores approaching 1 indicate that the document is highly related to the topic.

The summation of these node-level relevance scores across all connected dyads effectively assigns weights to citations, indicating whether each edge in the network is relevant to the topic (i.e., policing and life-course research):

$W_r(E_{i,j})=cos(document_i,topic)+cos(document_j,topic)$

Linear combination is unnecessary at this stage because we have no reason to assume that the topical relevance of the citing article is any more or less important than the topical relevance of the cited article. In other words, simple summation is sufficient. Conveniently, the dyadic summation of nodal cosine similarity to a given topic strengthens edges connecting relevant (+) nodes, diminishes edges connecting irrelevant (-) nodes, and ‘cancels out’ edges connecting relevant (+) to irrelevant (-) nodes (see Figure 2b-2c).

Next, borrowing the intuition of activation functions from the neural network literature, we take the rectified linear unit (ReLU) of each edge in the network. This effectively ‘switches off’ all citations with little or no topical relevance, when $W_r(E_{i,j})\leq\alpha$, while retaining all relevance scores above a certain threshold as edge weights:

$W_r(E_{i,j})^+=\frac{(W_r(E_{i,j})-\alpha)+|(W_r(E_{i,j})-\alpha)|}{2}$

The $\alpha$ hyperparameter allows the variable specification of a minimum required edge weight to be considered relevant to the focal research topic. Conveniently, cosine similarity is already centered on 0, so we set $\alpha$ to 0, activating all non-negative edges, $W_r(E_{i,j})>0$.[6] Increasing $\alpha$ serves to prune additional irrelevant edges from the citation network and allows the application of WAMPA to edge weights that are not typically centered on 0.

This produces $G(N,E,W_r)$, which we refer to as the ‘weight activated citation network’, where $N$ is a set of nodes that represent publications, $\{n_1,n_2,...,n_k\}$, $E$ is a set of acyclic, activated edges that represent topically-relevant citations, $\{\{i,j\}:i,j\in{N},i\neq{j})\}$, and $W_r$ is a weighted adjacency matrix describing the topical relevance of each edge where $W_r(E_{i,j})>0$ is required for $\{i,j\}\in{E}$. Notably, the activation function is applied prior to the calculation of path traversal weights (i.e., SPLC). Consequently, only topically relevant, activated paths are available for traversal, reducing the likelihood of semantic drift when approaching sources and sinks.

$W_c(E_{i,j})=\gamma\bigg(\frac{W_s(E_{i,j})-min(W_s)}{max(W_s)-min(W_s)}\bigg)+(1-\gamma)\bigg(\frac{W_t(E_{i,j})-min(W_t)}{max(W_t)-min(W_t)}\bigg)$

Finally, we employ the SPLC algorithm to calculate traversal weights for the weight activated citation network. This produces a second weighted adjacency matrix, $W_t$, representing the number of times each activated edge is traversed by paths from source to sink. We take the linear combination of the normalized relevance and traversal weights as the network weights and perform a key route search. The gamma ($\gamma$ ) hyperparameter varies between 0 and 1 and prioritizes either traversal or relevance weighting, respectively. We examine main paths derived of several different levels of $\gamma$. It should be noted that Chen et al. (2022) normalized and combined the edge weights during the main path search, ostensibly standardizing the edge weights within-path. However, we opt to normalize across the complete weight matrices. This is because, while sMPA is intended to ensure topical coherency within extracted main paths, WAMPA is intended to use relevance weights to compare topically-relevant main paths – normalizing the weights within each path would undermine this goal.

Results

Key routes through Criminology and Criminal Justice

Figure 2 presents the top 100 key routes through the criminology and criminal justice citation network, prior to weight activation and edge reweighting. Combined, these 100 key routes represent the ‘core’ of the field from 2001 until 2020. Of the 91 publications on these key routes, 30 were classified as “life-course” publications by top2vec (~33%), while none were classified as “policing” publications. DeLisi & Piquero (2011) was the most between central publication on the main path ($C_{betweenness}=973.92$), a majority of ‘shortest paths’ through the field can trace their lineage through this review. Baglivio & Wolff (2021) and Fox et al. (2015) were the most out- and in-degree central publications, respectively. Fox et al. (2015) was cited by a total of 15 other publications on the main path (16.5%), while Baglivio & Wolff (2021) cited 11 other publications on the main path (12.1%). Baglivio et al. (2015) was the most eigenvector central publication. It is clear from these publications, and the adjacent publications in the citation network, that the main path of criminology shifted focus to adverse childhood experiences (ACEs) soon after 2011. The relative late appearance of these network central publications (2011 – 2020) is likely to due to a recency bias incurred by an increasing publication rate in Criminology, resulting in greater substantive diversity and opportunity for citation among recent publications. The earliest publications on the main path consist of Turner & Piquero (2002), Piquero & Buka (2002), and Piquero et al. (2002), all of which empirically support the widespread but anecdotal understanding that modern criminology is rooted in life-course theory.

Supporting the high face validity of our MPA results, we noted that the average citation count for main path publications was 51 times higher than the average across all publications (Main Path: $\bar{x}=115.55$, $sd=177.07$; Population: $\mu=11.9$, $\sigma=32.09$; $t=9.55$, $p<2.2e^{-16}$). The most cited publication in our data (Bernstein et al., 2003; $N_{cites}\geq{3170}$) did not fall on this main path because it was only cited by 101 publications that fit within the scope of criminology, as defined by dimensions.ai, placing it toward the periphery of the discipline. The publication cited most by criminological scholarship (Sunshine & Tyler, 2003; $N_{cites}\geq{1622}$) did not fall on this main path either, discussing policing and procedural justice rather than life-course theory. However, more appropriately, Sunshine & Tyler (2003) appears on the policing main path visualized by Figure 3.

Life-course and policing divide

Figure 3 presents the top 100 key routes through criminology and criminal justice citation network, combined with the top 100 key routes through both life-course and policing weight-activated citation networks. The main paths for life-course and policing research were identified by deactivating topically relevant edges ($cos\leq0$ ) prior to performing key route searches. This network represents the unionized main paths of criminology and criminal justice, life-course criminology, and policing scholarship between 2001 and 2020. The most in-degree central publication along the policing main path, and in the entire network, was Sunshine & Tyler (2003) ( $C_{indegree}=49$), with Wolfe et al. (2016) as the most out-degree central ($C_{outdegree}=20$ ), firmly affixing the substantive focus of this main path to procedural justice and police legitimacy. Consistent with the previous observation that the main path of criminology intersects with life-course research, the most in-degree central publication along the life-course path is Fox et al. (2015) ($C_{indegree}=34$ ), the most out-degree central is DeLisi & Piquero (2011) ($C_{outdegree}=31$ ).

Consistent with its purpose as a measure of brokerage, one of the most between central publications ( $C_{betweenness}=1208.65$) was Wolfe et al. (2016), the only publication to cite across the policing and life-course gulf within a 3-year period (accounting for online first publishing). However, the cited article, Jennings et al. (2012), was not classified as life-course research by our model, included in the network by virtue of its position on the main path of criminology and criminal justice rather than its relevance to life-course or policing. Further, Jennings et al. (2012) is a review of victim-offender overlap research, implying that the communication lag between the two subfields is even greater than this connection would initially imply – review articles are inherently retrospective. In the complete graph, where no edges are removed, only 2 publications on the main path of life-course criminology received citation from the policing main path, Pogarsky (2004) and Nagin & Pogarsky (2001), the latter of which is likely a misclassification by the topic model, focusing more on deterrence theory[7]. The earliest policing publication to cite these publications was Ratcliffe et al. (2011), implying a 7-year lag in communication between policing and life-course scholarship (Ratcliffe et al., 2011 → Pogarsky, 2004). On the other hand, only 1 publication on the main path of policing received citation from the life-course main path (DeLisi & Vaughn, 2014 → Mastrofski et al., 2002), implying that this limited exchange of ideas and siloing goes both ways. One especially important thing to note is that all these cross-path citations are received by Criminology (flagship) publications.

Figure 4 presents the top 100 key routes through criminology and criminal justice citation network, combined with the top 100 key routes through life-course and policing weight-activated citation networks, following the same procedures as Figure 3. However, the citation networks’ traversal weights were adjusted via linear combination with relevance weights. For each of the life-course and policing weight activated citation networks, we set gamma ($\gamma$) to the earliest distributional plateau of the maximum linear combination of relevance and traversal weights ($\gamma_{life-course}=0.34$, $\gamma_{policing}=0.65$, see Figure S2). We take this to represent the earliest point that relevance weights start to supercede traversal weights, ensuring that the main paths will better represent the topics of interest – recognizing that the previous policing main path excluded much of the work on policing strategies and practices. Thus, the main paths presented in this network further prioritize the relevance of each publication and citation to the focal topics of life-course and policing, presenting research that better reflects scholars’ appraisal of the literature. Notably, regardless of how high we set the gamma hyperparameter, we could not find any main paths with more than one citation that ‘bridged’ life-course and policing research within a 3-year period, implying limited active communication between the nuclei of these subfields.

Weisburd & Eck (2004) and Braga & Bond (2008) were the two policing publications with the greatest in-degree ($C_{indegree}=\{36,33\}$, respectively). DeLisi & Piquero (2011) is, again, the life-course publication with the greatest betweenness, in-degree, and out-degree centrality ($C_{betweenness}=3385.2$; $C_{indegree}=32$; $C_{outdegree}=35$), appearing to serve as a turning point in both criminology and life-course theory. Unsurprisingly, the policing publications with the greatest out-degree are similarly seminal review articles: Braga et al. (2019a) and Braga et al. (2019b). Eigenvector centrality was concentrated among policing publications, including Braga & Bond (2008) ($C_{eigenvector}=1$), Braga et al. (2019b) ($C_{eigenvector}=0.98$), and Weisburd & Eck (2004) ($C_{eigenvector}=0.91$). This suggests a tendency for policing publications along the main paths to funnel towards a relatively small subset of prestigious publications when compared to life-course research.

On the other hand, betweenness centrality was concentrated among life-course publications, including Jennings & Reingle (2012) ($C_{betweenness}=2041.2$), Cullen (2011) ($C_{betweenness}=1732.2$), and Baglivio et al. (2015) ($C_{betweenness}=1432.8$). Cullen (2011) was the only direct bridge from policing to life-course, cited by Weisburd (2015). In the complete graph, this intersection between policing and life-course scholarship was part of a transitive triad where Weisburd (2015) → Cullen (2011) → Lum et al. (2011), and Weisburd (2015) → Lum et al. (2011); representing the debate on the future of criminology vis-à-vis adolescence-limited research. The other bridges are similarly lacking in the effective, meaningful exchange of substantive ideas and concepts. Crow et al. (2023), which was classified as life-course by virtue of a reference to biological scholarship in the context of unpublished null findings in criminology and criminal justice, cited Rothstein (2008) which was similarly misclassified due to a persistent association between the words “meta-analysis” and “policing.” Schnell & McManus (2022) cited Skardhamar (2010) in reference to temporal specification in (hot spots) policing research, with a specific emphasis on how policing interventions require shorter temporal windows than socio-behavioral research. The only exception is Petkovsek et al.'s (2016) citation of Katz et al. (2001), noting that Katz et al. (2001) serves as evidence that an arrest → gang membership causal path is relatively rare compared to its inverse. Nevertheless, this is very much ‘the exception that proves the rule’ that salient exchanges between the main paths of policing and life-course scholarship are predominantly centered around either seminal works that are difficult to ignore (i.e., Criminology publications) or philosophical and methodological disagreements.

Discussion

This paper examined cross-disciplinary communication between life-course criminology and policing research. We applied a modified, weight activated main path analysis (WAMPA) to criminology abstracts, identifying important contributions to the field and assessing the overlap between life-course and policing scholarship. Our findings revealed minimal cross-citation between the most influential works of life-course and policing research. This dissonance suggests that life-course criminology and policing scholarship have evolved along separate paths, each developing its unique knowledge with limited cross-pollination. The few instances of cross-citation observed in the data reflected the citation of major, flagship publications – which are inherently salient to the entire field, regardless of their relevance or complementarity – and fundamental (philosophical, theoretical, or methodological) disagreements. Scarce communication between subfields has significant implications for the entirety of the discipline. As a specialty area, criminology aims to understand crime and criminal behavior within a broader social context, integrating knowledge from a variety of other fields (Cressey, 1979; Leahey & Moody, 2014). A lack of integration, or at the very least lagged communication between subfields (by our estimates, over 4 years) reveals a clear lack of self-awareness as a discipline – despite repeated, prominent reminders that, as criminologists, we have often failed to critically reflect on the state and development of the field (see Barnes et al., 2014a; Barnes et al., 2014b; Garland, 1985, 2011; Laub, 2004). The absence of an intersection between prominent subfields like policing and life-course criminology contradicts the integrative and multidisciplinary nature of criminology, namely the integration of knowledge from various sciences to understand human behavior as it pertains to the criminal justice system (Petersilia & Sampson, 2018).

The isolation of subfields can impede the development of a holistic understanding of criminal behavior and its prevention (Farrington & Welsh, 2008). There is a myriad of credible challenges that impede collaboration between different specialties, including difficulties in synthesizing or agreeing on meritorious ideas (Lamont, 2010; Lamont et al., 2006), a lack of personal and professional motivation and support (Siedlok & Hibbert, 2014), and the high risk (high reward) nature of cross-(sub)disciplinary work (Leahey, 2016). Yet subfield integration can promote innovation, wherein research spanning multiple subfields can prompt creative and diverse solutions to real world problems, with each subfield inspiring (or forcing through debate and disagreement) ‘fresh thinking’ in the others (Leahey & Moody, 2014; Uzzi & Spiro, 2005). While the sub-disciplinary siloing we have observed here is not uncommon among the social sciences, there is attendant evidence that integrated works tend to have a more significant impact, generating more citations after controlling for the relative productivity of each subfield (Leahey & Moody, 2014). Moreover, specialization is one of the most cited drivers of collaboration in science (Leahey, 2016), recognizing that the increasing complexity and scope of collected knowledge (i.e., ‘information overload’) has made it untenable for any single scholar or group of scholars to address increasingly complex issues (Evans, 2008; Landhuis, 2016). Given that the complex nature of the criminal justice and legal system begets specialization (courts, corrections, law enforcement, etc.), collaboration between these specializations might become a necessary step for the advancement of the field.

Further, there are significant benefits to facilitating or encouraging more collaboration between life-course and policing specifically. A lack of integrated research undermines the potential for developing evidence-based policing practices that address the complexities of criminal behavior across the life-course (Bacon, 2022; King, 2009). Law enforcement agencies, corrections, and community-based organizations operate within the interconnected landscape of the criminal justice system. The absence of clear collaboration in research pertaining to these different facets of criminality and the criminal justice system may result in fragmented practices and policies that do not consider their interaction. Echoing our earlier example, police frequently interact with individuals of all ages, from early childhood in educational settings to the ever-aging prison population (Tapp & Davis, 2022). The absence of state-level minimum age laws for juvenile court jurisdiction (Abrams et al., 2019), the growing presence of school resource officers throughout childhood and adolescence (Musu-Gillette et al., 2018), and a relative dearth of recent research on specialized police services for the elderly (Zevitz & Rettammel, 1990) could reflect the failure of policing research to integrate the developmental perspective, and vice versa.

Indeed, the life-course perspective underscores the importance of developmentally informed law enforcement practices. Adolescents and young adults undergo significant cognitive, emotional, and social changes that influence their behavior and decision-making processes (Sampson & Laub, 2015). They are more prone to impulsive behavior and are less capable of fully understanding the long-term consequences of their actions (Steinberg, 2010). Given the potential of police contact to disrupt critical developmental processes (Loeber & Farrington, 2000), it stands to reason that tailored law enforcement practices would better serve adolescents and young adults. In a similar vein, life-course and policing scholars could collaboratively develop evidence-based policies that clearly define the role of school resource officers in a way that supports developmental needs (Musu-Gillette et al., 2018). Put simply, life-course and policing scholars could work together to propose and test policies and practices that prioritize the educational and emotional development of adolescents while minimizing unnecessary justice system contact. If effective, this collaborative effort could simultaneously mitigate the risk of escalating police contacts and foster positive outcomes by aligning interventions with developmental needs.

In addition, integrating life-course research with policing is crucial for addressing disparities in how age, culpability, and susceptibility to harm are perceived across different demographic groups. Some law enforcement officers perceive age, culpability, and susceptibility to harm differently based on race and gender, which can lead to disparities in treatment (Goff et al., 2014; Perillo et al., 2023). Research suggests law enforcement officers perceive Black children as older and more culpable than their white peers (Goff et al., 2014), which can result in harsher treatment and increased involvement with the justice system. Similarly, Perillo et al. (2023) found evidence that gendered perceptions may affect how law enforcement officers assess culpability and vulnerability, contributing to differential treatment. Incorporating a life-course perspective into training and policy development could help address these biases. By raising awareness of how race and gender can influence perceptions and treatment, law enforcement agencies can develop more equitable practices and effective training programs. 

While recognizing the importance of this work to research and policy, there are several limitations. Like arrest data, publication data and co-citation networks inevitably suffer from only capturing formal, published communication that successfully traversed the peer-review process; a “dark figure of collaboration” looms large. Academics formally collaborate on many other activities – grants, committee assignments, conference service, etc. (Leahey, 2016; Leone Sciabolazza et al., 2020) – and informally communicate and interact in any number of additional ways. Consequently, it is likely that life-course and policing scholars do interact and influence each other to a greater extent than is implied by this analysis. While many of these interactions are not public facing and would not influence the development of an academic subfield without formal acknowledgement through publication and citation, it is worth noting that studies attempting to integrate multiple subfields might be languishing in the “file drawer”, never to see the light of day (Crow et al., 2023). Indeed, it has long been recognized that unconventional research that attempts to bridge disparate (sub)disciplines can be viewed negatively by reviewers and editors (Birnbaum, 1981; Mansilla, 2006). This could imply that (1) editors and reviewers might be unwilling to endorse research that bridges life-course and policing scholarship, or (2) previous research studying intersections between life-course and policing scholarship may have been fruitless. The former does not necessarily undermine our findings, instead highlighting another barrier that must be considered when facilitating and encouraging inter-subfield scholarship in criminology and criminal justice. Whereas the latter seems increasingly unlikely given the myriad ways that the life-course stage of a suspect or victim intersects with police policy and practice (Abrams et al., 2020; Loeber & Farrington, 2000; Novak & De Francisco Lopes, 2023). Regardless, the potential to obfuscate innovative inter-subfield research that does not fit neatly into the field’s journals underscores criminology’s need to reconcile with the “missing null” and encourage open science practices (Crow et al., 2023; Greenspan et al., 2024; Pridemore et al., 2018).

Additionally, the devising and application of WAMPA inherently narrowed the focus of this project to the ‘core’ of each subfield. This inherently omits more peripheral interactions between policing and life-course scholars. Indeed, research on the school-to-prison pipeline (Skiba et al., 2014) and the victim-offender overlap (Jennings et al., 2010) could be viewed as examples of successful intersections that would not necessarily appear in the ‘core’ of any one subfield of criminology. However, these topics could very well serve as the exception to the rule. We explicitly note that Jennings et al. (2012), a review of literature on the victim-offender overlap, appears on the main citation path, clearly demonstrating that research at the intersection of different topics in criminology are not inherently relegated to the periphery of the field. Moreover, as noted in our literature review, there are relatively few explicit integrations of life-course and policing scholarship (Bosick et al., 2012; Howes & Goodman-Delahunty, 2014; King, 2009; Schuck, 2013). This suggests that peripheral cross-citations are similarly limited. It should be further noted that topics occupying a conceptual intersection between criminological subfields are not necessarily the result of cross-pollination. Despite being relevant to corrections scholarship, school-to-prison pipeline research is predominantly the result of cross-disciplinary collaboration between researchers in criminology, social work, and education – not life-course and corrections (Mallett, 2016). Finally, briefly recognizing, discussing, and citing the relevance of important work from other subfields is a low bar, and does not face the same barriers as co-authorship and collaboration (see Leahey, 2016). Nevertheless, by introducing WAMPA, we showcase a novel approach to guiding MPA by integrating unsupervised machine learning methods. This model can help researchers better understand the development of a subfield without entirely divorcing that subfield from its wider disciplinary context.

Ultimately, the limited, lagged communication between two important subfields of criminology underscores a need for a science of criminology (Smith et al., 2023a), following in the footsteps of the science of science (Fortunato et al., 2018). A lack of communication between subfields could obfuscate redundant efforts and missed opportunities for innovation. But we cannot identify these redundancies and inefficiencies without taking stock of the field – a goal that is beyond the scope of systematic reviews and meta-analyses (Landhuis, 2016; Turanovic & Pratt, 2021). Strategically encouraging research that bridges complementary subfields, organizing inter-subfield symposia, and funding collaborative research projects could help mitigate sub-disciplinary siloing and foster a more integrated approach to criminological research. By creating platforms for dialogue and collaboration, we might leverage the uniquely diverse expertise of criminologists and address increasingly complex criminal justice and legal issues.

Bibliography

Abrams, L. S., Barnert, E. S., Mizel, M. L., Bedros, A., Webster, E., & Bryan, I. (2020). When Is a Child Too Young for Juvenile Court? A Comparative Case Study of State Law and Implementation in Six Major Metropolitan Areas. Crime & Delinquency, 66(2), 219–249. https://doi.org/10.1177/0011128719839356

Abrams, L. S., Barnert, E. S., Mizel, M. L., Bryan, I., Lim, L., Bedros, A., Soung, P., & Harris, M. (2019). Is a Minimum Age of Juvenile Court Jurisdiction a Necessary Protection? A Case Study in the State of California. Crime & Delinquency, 65(14), 1976–1996. https://doi.org/10.1177/0011128718770817

Agnew, R., & Brezina, T. (2019). General Strain Theory. In M. D. Krohn, N. Hendrix, G. Penly Hall, & A. J. Lizotte (Eds.), Handbook on Crime and Deviance (pp. 145–160). Springer International Publishing. https://doi.org/10.1007/978-3-030-20779-3_8

Andrews, D. A., Bonta, J., & Wormith, J. S. (2011). The Risk-Need-Responsivity (RNR) Model: Does Adding the Good Lives Model Contribute to Effective Crime Prevention? Criminal Justice and Behavior, 38(7), 735–755. https://doi.org/10.1177/0093854811406356

Angelov, D. (2020). Top2Vec: Distributed Representations of Topics (arXiv:2008.09470). arXiv. https://doi.org/10.48550/arXiv.2008.09470

Australian Bureau of Statistics. (2020). Australian and New Zealand Standard Research Classification (ANZSRC). https://www.abs.gov.au/statistics/classifications/australian-and-new-zealand-standard-research-classification-anzsrc/latest-release

Bacon, M. (2022). Desistance from criminalisation: Police culture and new directions in drugs policing. Policing and Society, 32(4), 522–539. https://doi.org/10.1080/10439463.2021.1920587

Baglivio, M. T., & Wolff, K. T. (2021). Positive Childhood Experiences (PCE): Cumulative Resiliency in the Face of Adverse Childhood Experiences. Youth Violence and Juvenile Justice, 19(2), 139–162. https://doi.org/10.1177/1541204020972487

Baglivio, M. T., Wolff, K. T., Piquero, A. R., & Epps, N. (2015). The Relationship between Adverse Childhood Experiences (ACE) and Juvenile Offending Trajectories in a Juvenile Offender Sample. Journal of Criminal Justice, 43(3), 229–241. https://doi.org/10.1016/j.jcrimjus.2015.04.012

Barnes, J. C., Boutwell, B. B., Beaver, K. M., Gibson, C. L., & Wright, J. P. (2014a). On the consequences of ignoring genetic influences in criminological research. Journal of Criminal Justice, 42(6), 471-482. https://doi.org/10.1016/j.jcrimjus.2014.08.003

Barnes, J. C., Wright, J. P., Boutwell, B. B., Schwartz, J. A., Connolly, E. J., Nedelec, J. L., & Beaver, K. M. (2014b). Demonstrating the validity of twin research in criminology. Criminology, 52(4), 588-626. https://psycnet.apa.org/doi/10.1111/1745-9125.12049

Beckley, A. L., Caspi, A., Arseneault, L., Barnes, J. C., Fisher, H. L., Harrington, H., Houts, R., Morgan, N., Odgers, C. L., Wertz, J., & Moffitt, T. E. (2018). The Developmental Nature of the Victim-Offender Overlap. Journal of Developmental and Life-Course Criminology, 4(1), 24–49. https://doi.org/10.1007/s40865-017-0068-3

Berg, M. T., & Schreck, C. J. (2022). The Meaning of the Victim–Offender Overlap for Criminological Theory and Crime Prevention Policy. Annual Review of Criminology, 5(Volume 5, 2022), 277–297. https://doi.org/10.1146/annurev-criminol-030920-120724

Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., Stokes, J., Handelsman, L., Medrano, M., Desmond, D., & Zule, W. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse & Neglect, 27(2), 169–190. https://doi.org/10.1016/S0145-2134(02)00541-0

Biancani, S., & McFarland, D. A. (2014). The Semiformal Organization. Organization Science, 25(5), 1306–1324.

Birnbaum, P. H. (1981). Integration and Specialization in Academic Research. Academy of Management Journal, 24(3), 487–503. https://doi.org/10.5465/255570

Bosick, S. J., Rennison, C. M., Gover, A. R., & Dodge, M. (2012). Reporting violence to the police: Predictors through the life course. Journal of Criminal Justice, 40(6), 441–451. https://doi.org/10.1016/j.jcrimjus.2012.05.001

Braga, A. A., & Bond, B. J. (2008). Policing Crime and Disorder Hot Spots: A Randomized Controlled Trial. Criminology, 46(3), 577–607. https://doi.org/10.1111/j.1745-9125.2008.00124.x

Braga, A. A., Turchan, B., Papachristos, A. V., & Hureau, D. M. (2019a). Hot spots policing of small geographic areas effects on crime. Campbell Systematic Reviews, 15(3), e1046. https://doi.org/10.1002/cl2.1046

Braga, A. A., Turchan, B. S., Papachristos, A. V., & Hureau, D. M. (2019b). Hot spots policing and crime reduction: An update of an ongoing systematic review and meta-analysis. Journal of Experimental Criminology, 15(3), 289–311. https://doi.org/10.1007/s11292-019-09372-3

Brint, S. (2005). Creating the Future: “New Directions” in American Research Universities. Minerva, 43(1), 23–50.

Brogan, L., Haney-Caron, E., NeMoyer, A., & DeMatteo, D. (2015). Applying the Risk-Needs-Responsivity (RNR) Model to Juvenile Justice. Criminal Justice Review, 40(3), 277–302. https://doi.org/10.1177/0734016814567312

Bucerius, S. M., Jones, D. J., Kohl, A., & Haggerty, K. D. (2021). Addressing the Victim–Offender Overlap: Advancing Evidence-Based Research to Better Service Criminally Involved People with Victimization Histories. Victims & Offenders, 16(1), 148–163. https://doi.org/10.1080/15564886.2020.1787283

Campello, R. J. G. B., Moulavi, D., & Sander, J. (2013). Density-Based Clustering Based on Hierarchical Density Estimates. In J. Pei, V. S. Tseng, L. Cao, H. Motoda, & G. Xu (Eds.), Advances in Knowledge Discovery and Data Mining (pp. 160–172). Springer. https://doi.org/10.1007/978-3-642-37456-2_14

Chen, L., Xu, S., Zhu, L., Zhang, J., Xu, H., & Yang, G. (2022). A semantic main path analysis method to identify multiple developmental trajectories. Journal of Informetrics, 16(2), 101281. https://doi.org/10.1016/j.joi.2022.101281

Cohen, L. E., & Felson, M. (1979). Social Change and Crime Rate Trends: A Routine Activity Approach. American Sociological Review, 44(4), 588–608. https://doi.org/10.2307/2094589

Connolly, E. J., & Beaver, K. M. (2015). Assessing the salience of gene–environment interplay in the development of anger, family conflict, and physical violence: A biosocial test of General Strain Theory. Journal of Criminal Justice, 43(6), 487–497. https://doi.org/10.1016/j.jcrimjus.2015.11.001

Counts, J., Randall, K. N., Ryan, J. B., & Katsiyannis, A. (2018). School Resource Officers in Public Schools: A National Review. Education and Treatment of Children, 41(4), 405–430.

Cressey, D. R. (1979). Fifty Years of Criminology: From Sociological Theory to Political Control. Pacific Sociological Review, 22(4), 457–480. https://doi.org/10.2307/1388842

Crow, M. S., Smykla, J. O., O’Brien, H., Cerna, T., Johnson, A., Pisaris, S., Suarez, M., & Wilder, J. (2023). What’s in Your File Drawer? The Case of the Missing Null in Criminology and Criminal Justice. Crime & Delinquency, 69(12), 2574–2594. https://doi.org/10.1177/00111287221090959

Csardi, G., & Nepusz, T. (2005). The igraph software package for complex network research. InterJournal, Complex Systems, 1695.

Cullen, F. T. (2011). Beyond Adolescence-Limited Criminology: Choosing Our Future—the American Society of Criminology 2010 Sutherland Address. Criminology, 49(2), 287–330. https://doi.org/10.1111/j.1745-9125.2011.00224.x

Davies, M., & Ruda, S. (2023). Reflections on applying behavioural insights to crime: A guide for behavioural scientists and criminologists in search of policy unicorns. Behavioural Public Policy, 7(3), 744–757. https://doi.org/10.1017/bpp.2023.13

DeLisi, M., & Piquero, A. R. (2011). New frontiers in criminal careers research, 2000–2011: A state-of-the-art review. Journal of Criminal Justice, 39(4), 289–301. https://doi.org/10.1016/j.jcrimjus.2011.05.001

DeLisi, M., & Vaughn, M. G. (2014). Foundation for a temperament-based theory of antisocial behavior and criminal justice system involvement. Journal of Criminal Justice, 42(1), 10–25. https://doi.org/10.1016/j.jcrimjus.2013.11.001

Donnermeyer, J., & DeKeseredy, W. (2013). Rural Criminology. Routledge. https://doi.org/10.4324/9780203094518

Elder, G. H. (1985). Perspectives on the life course. In G. H. Elder (Ed.), Life course dynamics (pp. 23–49). Cornell University Press.

Evans, J. A. (2008). Electronic Publication and the Narrowing of Science and Scholarship. Science, 321(5887), 395–399. https://doi.org/10.1126/science.1150473

Farrington, D. P. (1986). Age and Crime. Crime and Justice, 7, 189–250. https://doi.org/10.1086/449114

Farrington, D. P., & Welsh, B. C. (2008). Saving Children from a Life of Crime: Early Risk Factors and Effective Interventions. Oxford University Press.

Fortunato, S., Bergstrom, C. T., Börner, K., Evans, J. A., Helbing, D., Milojević, S., Petersen, A. M., Radicchi, F., Sinatra, R., Uzzi, B., Vespignani, A., Waltman, L., Wang, D., & Barabási, A.-L. (2018). Science of science. Science, 359(6379), eaao0185. https://doi.org/10.1126/science.aao0185

Fox, B. H., Perez, N., Cass, E., Baglivio, M. T., & Epps, N. (2015). Trauma changes everything: Examining the relationship between adverse childhood experiences and serious, violent and chronic juvenile offenders. Child Abuse & Neglect, 46, 163–173. https://doi.org/10.1016/j.chiabu.2015.01.011

Garfield, E. (1979). Citation Indexing—Its Theory and Application in Science, Technology, and Humanities. Wiley.

Garland, D. (1985). THE CRIMINAL AND HIS SCIENCE: A Critical Account of the Formution of Criminology at the End of the Nineteenth century. The British Journal of Criminology, 25(2), 109–137. https://doi.org/10.1093/oxfordjournals.bjc.a047507

Garland, D. (2011). Criminology’s Place in the Academic Field. In What is Criminology? (pp. 298–317). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199571826.003.0021

Goff, P. A., Jackson, M. C., Di Leone, B. A. L., Culotta, C. M., & DiTomasso, N. A. (2014). The essence of innocence: Consequences of dehumanizing Black children. Journal of Personality and Social Psychology, 106(4), 526–545. https://doi.org/10.1037/a0035663

Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime (pp. xvi, 297). Stanford University Press.

Granot, Y., & Tyler, T. R. (2019). Adolescent cognition and procedural justice: Broadening the impact of research findings on policy and practice. Social and Personality Psychology Compass, 13(10), e12503. https://doi.org/10.1111/spc3.12503

Greenspan, R. L., Baggett, L., & B. Boutwell, B. (2024). Open science practices in criminology and criminal justice journals. Journal of Experimental Criminology. https://doi.org/10.1007/s11292-024-09640-x

Holt, T. V., Johnson, J. C., Moates, S., & Carley, K. M. (2016). The Role of Datasets on Scientific Influence within Conflict Research. PLOS ONE, 11(4), e0154148. https://doi.org/10.1371/journal.pone.0154148

Honnibal, M., & Montani, I. (2017). spaCy 2: Natural language understanding with Bloom embeddings, convolutional neural networks and incremental parsing. Sentometrics Research. https://sentometrics-research.com/publication/72/

Howes, L. M., & Goodman-Delahunty, J. (2014). Life Course Research Design: Exploring Career Change Experiences of Former School Teachers and Police Officers. Journal of Career Development, 41(1), 62–84. https://doi.org/10.1177/0894845312474370

Huff, J. (2021). Understanding police decisions to arrest: The impact of situational, officer, and neighborhood characteristics on police discretion. Journal of Criminal Justice, 75, 101829. https://doi.org/10.1016/j.jcrimjus.2021.101829

Hummon, N. P., & Dereian, P. (1989). Connectivity in a citation network: The development of DNA theory. Social Networks, 11(1), 39–63. https://doi.org/10.1016/0378-8733(89)90017-8

Jeffery, C. R. (1978). Criminology as an Interdisciplinary Behavioral Science. Criminology, 16, 149.

Jennings, W. G., Higgins, G. E., Tewksbury, R., Gover, A. R., & Piquero, A. R. (2010). A Longitudinal Assessment of the Victim-Offender Overlap. Journal of Interpersonal Violence, 25(12), 2147–2174. https://doi.org/10.1177/0886260509354888

Jennings, W. G., Piquero, A. R., & Reingle, J. M. (2012). On the overlap between victimization and offending: A review of the literature. Aggression and Violent Behavior, 17(1), 16–26. https://doi.org/10.1016/j.avb.2011.09.003

Jennings, W. G., & Reingle, J. M. (2012). On the number and shape of developmental/life-course violence, aggression, and delinquency trajectories: A state-of-the-art review. Journal of Criminal Justice, 40(6), 472–489. https://doi.org/10.1016/j.jcrimjus.2012.07.001

Katz, C. M., Webb, V. J., & Schaefer, D. R. (2001). An assessment of the impact of quality-of-life policing on crime and disorder. Justice Quarterly, 18(4), 825–876. https://doi.org/10.1080/07418820100095111

King, W. R. (2009). Toward a Life-Course Perspective of Police Organizations. Journal of Research in Crime and Delinquency, 46(2), 213–244. https://doi.org/10.1177/0022427808330874

Krohn, M. D., & Eassey, J. M. (2014). Integrated theories of crime. In J. M. Miller (Ed.), The encyclopedia of theoretical criminology (pp. 1–6). Wiley.

Kurlychek, M. C., & Johnson, B. D. (2019). Cumulative Disadvantage in the American Criminal Justice System. Annual Review of Criminology, 2(Volume 2, 2019), 291–319. https://doi.org/10.1146/annurev-criminol-011518-024815

Lamont, M. (2010). How Professors Think: Inside the Curious World of Academic Judgement. Harvard University Press. https://www.hup.harvard.edu/books/9780674057333

Lamont, M., Mallard, G., & Guetzkow, J. (2006). Beyond blind faith: Overcoming the obstacles to interdisciplinary evaluation. Research Evaluation, 15(1), 43–55. https://doi.org/10.3152/147154406781776002

Landhuis, E. (2016). Scientific literature: Information overload. Nature, 535(7612), Article 7612. https://doi.org/10.1038/nj7612-457a

Laub, J. H. (2004). The Life Course of Criminology in the United States: The American Society of Criminology 2003 Presidential Address. Criminology, 42(1), 1–26. https://doi.org/10.1111/j.1745-9125.2004.tb00511.x

Laub, J. H., & Sampson, R. J. (2020). Life-Course and Developmental Criminology: Looking Back, Moving Forward—ASC Division of Developmental and Life-Course Criminology Inaugural David P. Farrington Lecture, 2017. Journal of Developmental and Life-Course Criminology, 6(2), 158–171. https://doi.org/10.1007/s40865-019-00110-x

Leahey, E. (2016). From Sole Investigator to Team Scientist: Trends in the Practice and Study of Research Collaboration. Annual Review of Sociology, 42(Volume 42, 2016), 81–100. https://doi.org/10.1146/annurev-soc-081715-074219

Leahey, E., & Moody, J. (2014). Sociological Innovation through Subfield Integration. Social Currents, 1(3), 228–256. https://doi.org/10.1177/2329496514540131

Leone Sciabolazza, V., Vacca, R., & McCarty, C. (2020). Connecting the dots: Implementing and evaluating a network intervention to foster scientific collaboration and productivity. Social Networks, 61, 181–195. https://doi.org/10.1016/j.socnet.2019.11.003

Liebling, A. (2000). Prison Officers, Policing and the Use of Discretion. Theoretical Criminology, 4(3), 333–357. https://doi.org/10.1177/1362480600004003005

Liu, J. S., & Lu, L. Y. Y. (2012). An integrated approach for main path analysis: Development of the Hirsch index as an example. Journal of the American Society for Information Science and Technology, 63(3), 528–542. https://doi.org/10.1002/asi.21692

Loeber, R., & Farrington, D. P. (2000). Young children who commit crime: Epidemiology, developmental origins, risk factors, early interventions, and policy implications. Development and Psychopathology, 12(4), 737–762. https://doi.org/10.1017/s0954579400004107

Lum, C., Koper, C. S., & Telep, C. W. (2011). The Evidence-Based Policing Matrix. Journal of Experimental Criminology, 7(1), 3–26. https://doi.org/10.1007/s11292-010-9108-2

Mallett, C. A. (2016). The School-to-Prison Pipeline: A Critical Review of the Punitive Paradigm Shift. Child and Adolescent Social Work Journal, 33(1), 15–24. https://doi.org/10.1007/s10560-015-0397-1

Mansilla, V. B. (2006). Assessing expert interdisciplinary work at the frontier: An empirical exploration. Research Evaluation, 15(1), 17–29. https://doi.org/10.3152/147154406781776075

Mastrofski, S. D., Reisig, M. D., & McCLUSKEY, J. D. (2002). Police Disrespect Toward the Public: An Encounter-Based Analysis. Criminology, 40(3), 519–552. https://doi.org/10.1111/j.1745-9125.2002.tb00965.x

McGlynn-Wright, A., Crutchfield, R. D., Skinner, M. L., & Haggerty, K. P. (2022). The Usual, Racialized, Suspects: The Consequence of Police Contacts with Black and White Youth on Adult Arrest. Social Problems, 69(2), 299–315. https://doi.org/10.1093/socpro/spaa042

McInnes, L., Healy, J., & Melville, J. (2020). UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction (arXiv:1802.03426). arXiv. https://doi.org/10.48550/arXiv.1802.03426

Merton, R. K. (1973). The Sociology of Science: Theoretical and Empirical Investigations. University of Chicago Press.

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100(4), 674–701.

Murphy, K. (2015). Does procedural justice matter to youth? Comparing adults’ and youths’ willingness to collaborate with police. Policing and Society, 25(1), 53–76. https://doi.org/10.1080/10439463.2013.802786

Musu-Gillette, L., Zhang, A., Wang, K., Zhang, J., Kemp, J., Diliberti, M., & Oudekerk, B. (2018). Indicators of School Crime and Safety: 2017. National Center for Education Statistics. https://nces.ed.gov/pubs2018/2018036.pdf

Nagin, D. S., & Pogarsky, G. (2001). Integrating Celerity, Impulsivity, and Extralegal Sanction Threats into a Model of General Deterrence: Theory and Evidence. Criminology, 39(4), 865–892. https://doi.org/10.1111/j.1745-9125.2001.tb00943.x

Novak, A., Boutwell, B. B., & Smith, T. B. (2023). Taking the problem of colliders seriously in the study of crime: A research note. Journal of Experimental Criminology. https://doi.org/10.1007/s11292-023-09565-x

Novak, A., & De Francisco Lopes, V. (2023). What About the kids? A Multimethod Approach to Understanding Law Enforcement Policies Pertaining to the Arrest of Children in Florida. American Journal of Criminal Justice, 48(4), 899–920. https://doi.org/10.1007/s12103-023-09729-5

Novak, A., & De Francisco Lopes, V. (2024). Structuring Discretion: The Association Between Law Enforcement Policies and Arrests of Children in Florida. Crime & Delinquency, 70(6–7), 1615–1638. https://doi.org/10.1177/00111287221134046

Perillo, J. T., Sykes, R. B., Bennett, S. A., & Reardon, M. C. (2023). Examining the consequences of dehumanization and adultification in justification of police use of force against Black girls and boys. Law and Human Behavior, 47(1), 36–52. https://doi.org/10.1037/lhb0000521

Petersilia, J., & Sampson, R. J. (2018). Charting a Path Forward for Criminology. Annual Review of Criminology, 1(Volume 1, 2018), i–iv. https://doi.org/10.1146/annurev-cr-01-120717-100001

Petkovsek, M. A., Boutwell, B. B., Barnes, J. C., & Beaver, K. M. (2016). Moffitt’s Developmental Taxonomy and Gang Membership: An Alternative Test of the Snares Hypothesis. Youth Violence and Juvenile Justice, 14(4), 335–349. https://doi.org/10.1177/1541204015581389

Piquero, A. R., Brame, R., Mazerolle, P., & Haapanen, R. (2002). Crime in Emerging Adulthood. Criminology, 40(1), 137–170. https://doi.org/10.1111/j.1745-9125.2002.tb00952.x

Piquero, A. R., & Buka, S. L. (2002). Linking juvenile and adult patterns of criminal activity in the Providence cohort of the National Collaborative Perinatal Project. Journal of Criminal Justice, 30(4), 259–272. https://doi.org/10.1016/S0047-2352(02)00128-9

Pogarsky, G. (2004). Projected Offending and Contemporaneous Rule-Violation: Implications for Heterotypic Continuity. Criminology, 42(1), 111–136. https://doi.org/10.1111/j.1745-9125.2004.tb00515.x

Pridemore, W. A., Makel, M. C., & Plucker, J. A. (2018). Replication in Criminology and the Social Sciences. Annual Review of Criminology, 1(Volume 1, 2018), 19–38. https://doi.org/10.1146/annurev-criminol-032317-091849

Puzzanchera, C., Sladky, A., & Kang, W. (2023). Easy Access to Juvenile Court Statistics. https://www.ojjdp.gov/ojstatbb/ezajcs/

Ratcliffe, J. H., Taniguchi, T., Groff, E. R., & Wood, J. D. (2011). The Philadelphia Foot Patrol Experiment: A Randomized Controlled Trial of Police Patrol Effectiveness in Violent Crime Hotspots. Criminology, 49(3), 795–831. https://doi.org/10.1111/j.1745-9125.2011.00240.x

Reiss, A. J. (1971). The Police and the Public. Yale University Press.

Rothstein, H. R. (2008). Publication bias as a threat to the validity of meta-analytic results. Journal of Experimental Criminology, 4(1), 61–81. https://doi.org/10.1007/s11292-007-9046-9

Sampson, R. J., & Laub, J. H. (1990). Crime and Deviance over the Life Course: The Salience of Adult Social Bonds. American Sociological Review, 55(5), 609–627. https://doi.org/10.2307/2095859

Sampson, R. J., & Laub, J. H. (1993). Crime in the Making: Pathways and Turning Points Through Life. Crime & Delinquency, 39(3), 396–396. https://doi.org/10.1177/0011128793039003010

Sampson, R. J., & Laub, J. H. (2015). A Life-Course View of the Development of Crime. In Developmental and Life-course Criminological Theories. Routledge.

Schnell, C., & McManus, H. D. (2022). The Influence of Temporal Specification on the Identification of Crime Hot Spots for Program Evaluations: A Test of Longitudinal Stability in Crime Patterns. Journal of Quantitative Criminology, 38(1), 51–74. https://doi.org/10.1007/s10940-020-09483-5

Schuck, A. M. (2013). A Life-course Perspective on Adolescents’ Attitudes to Police: DARE, Delinquency, and Residential Segregation. Journal of Research in Crime and Delinquency, 50(4), 579–607. https://doi.org/10.1177/0022427813481977

Siedlok, F., & Hibbert, P. (2014). The Organization of Interdisciplinary Research: Modes, Drivers and Barriers. International Journal of Management Reviews, 16(2), 194–210. https://doi.org/10.1111/ijmr.12016

Skardhamar, T. (2010). Distinguishing Facts and Artifacts in Group-Based Modeling. Criminology, 48(1), 295–320. https://doi.org/10.1111/j.1745-9125.2010.00185.x

Skiba, R. J., Arredondo, M. I., & Williams, N. T. (2014). More Than a Metaphor: The Contribution of Exclusionary Discipline to a School-to-Prison Pipeline. Equity & Excellence in Education, 47(4), 546–564. https://doi.org/10.1080/10665684.2014.958965

Smith, T. B., Mao, R., Korotchenko, S., & Krohn, M. D. (2023a). Partners in Criminology: Machine Learning and Network Science Reveal Missed Opportunities and Inequalities in the Study of Crime. Journal of Quantitative Criminology. https://doi.org/10.1007/s10940-023-09574-z

Smith, T. B., Vacca, R., Mantegazza, L., & Capua, I. (2021). Natural language processing and network analysis provide novel insights on policy and scientific discourse around Sustainable Development Goals. Scientific Reports, 11(1), Article 1. https://doi.org/10.1038/s41598-021-01801-6

Smith, T. B., Vacca, R., Mantegazza, L., & Capua, I. (2023b). Discovering new pathways toward integration between health and sustainable development goals with natural language processing and network science. Globalization and Health, 19(1), 44. https://doi.org/10.1186/s12992-023-00943-8

Steinberg, L. (2010). A social neuroscience perspective on adolescent risk-taking. In Biosocial Theories of Crime. Routledge. Is this a book chapter/paper in a book? All I found with the same title is a journal paper: Steinberg L. (2008). A Social Neuroscience Perspective on Adolescent Risk-Taking. Developmental review : DR, 28(1), 78–106. https://doi.org/10.1016/j.dr.2007.08.002

Sunshine, J., & Tyler, T. R. (2003). The Role of Procedural Justice and Legitimacy in Shaping Public Support for Policing. Law & Society Review, 37(3), 513–547. https://doi.org/10.1111/1540-5893.3703002

Tapp, S. N., & Davis, E. J. (2022). Contacts Between Police and the Public, 2020. Bureau of Justice Statistics. https://bjs.ojp.gov/media/document/cbpp20.pdf

Taylor, A., Marcus, M., & Santorini, B. (2003). The Penn Treebank: An Overview. In A. Abeillé (Ed.), Treebanks: Building and Using Parsed Corpora (pp. 5–22). Springer Netherlands. https://doi.org/10.1007/978-94-010-0201-1_1

Theriot, M. T. (2016). The Impact of School Resource Officer Interaction on Students’ Feelings About School and School Police. Crime & Delinquency, 62(4), 446–469. https://doi.org/10.1177/0011128713503526

Theriot, M. T., & Cuellar, M. J. (2016). School resource officers and students’ rights. Contemporary Justice Review, 19(3), 363–379. https://doi.org/10.1080/10282580.2016.1181978

Turanovic, J. J., & Pratt, T. C. (2021). Meta-Analysis in Criminology and Criminal Justice: Challenging the Paradigm and Charting a New Path Forward. Justice Evaluation Journal, 4(1), 21–47. https://doi.org/10.1080/24751979.2020.1775107

Turner, M. G., & Piquero, A. R. (2002). The stability of self-control. Journal of Criminal Justice, 30(6), 457–471. https://doi.org/10.1016/S0047-2352(02)00169-1

Turney, K. (2022). The Mental Health Consequences of Vicarious Adolescent Police Exposure. Social Forces, 100(3), 1142–1169. https://doi.org/10.1093/sf/soab035

Uzzi, B., & Spiro, J. (2005). Collaboration and Creativity: The Small World Problem. American Journal of Sociology, 111(2), 447–504. https://doi.org/10.1086/432782

Walsh, A., & Ellis, L. (2006). Criminology: An Interdisciplinary Approach. SAGE.

Weatheritt, M. (2023). Innovations in Policing. Routledge. https://doi.org/10.4324/9781003371724

Weiler, S. C., & Cray, M. (2011). Police at School: A Brief History and Current Status of School Resource Officers. The Clearing House: A Journal of Educational Strategies, Issues and Ideas, 84(4), 160–163. https://doi.org/10.1080/00098655.2011.564986

Weisburd, D. (2015). The Law of Crime Concentration and the Criminology of Place. Criminology, 53(2), 133–157. https://doi.org/10.1111/1745-9125.12070

Weisburd, D., & Eck, J. E. (2004). What Can Police Do to Reduce Crime, Disorder, and Fear? The ANNALS of the American Academy of Political and Social Science, 593(1), 42–65. https://doi.org/10.1177/0002716203262548

Wiley, S. A., Slocum, L. A., & Esbensen, F.-A. (2013). The Unintended Consequences of Being Stopped or Arrested: An Exploration of the Labeling Mechanisms Through Which Police Contact Leads to Subsequent Delinquency. Criminology, 51(4), 927–966. https://doi.org/10.1111/1745-9125.12024

Wolfe, S. E., Nix, J., Kaminski, R., & Rojek, J. (2016). Is the Effect of Procedural Justice on Police Legitimacy Invariant? Testing the Generality of Procedural Justice and Competing Antecedents of Legitimacy. Journal of Quantitative Criminology, 32(2), 253–282. https://doi.org/10.1007/s10940-015-9263-8

Wu, L., Wang, D., & Evans, J. A. (2019). Large teams develop and small teams disrupt science and technology. Nature, 566, 378–382. https://doi.org/10.1038/s41586-019-0941-9

Zevitz, R. G., & Rettammel, R. J. (1990). Elderly attitudes about police service. Am. J. Police, 9, 25.

Supplemental Figures

[1] These data consisted predominantly of poorly maintained publication outlets and grey literature.

[2] The minimum requirement for dimensions.ai to assign a unique identifier to a given author following a process called “disambiguation”, the removal of ambiguity surrounding the identity of scholars in the data.

[3] Substantive POS include nouns (NN, NNP, NNS, NNPS), prepositions (IN), verbs (VB, VBD, VBG, VBN, VBP, VBZ), adverbs (RB, RBR, RBS, WRB), determiners (DT), adjectives (JJ, JJR, JJS), pronouns (PRP, WP), coordinating conjunctions (CC), cardinal numbers (CD), interjections (UH), and symbols (SYM).

[4] Policing keywords consisted of “police”, “policing”, “hot”, “spot”, “body”, “wear”; Life-course keywords consisted of “life”, “course”, “trajectory”, “onset”, “desistance”. We narrowed keywords to the smallest set that would accurately describe the field itself (i.e., “policing” and “life course”) as well as some of the most salient research in that subfield (i.e., “hot spots”, “body-worn cameras”, “trajectories”, “onset”, and “desistance”). This ensured that we cast the widest possible net or, more specifically, pinpoint the regions of the topic embedding space that contained most of the research related to these subfields. While some important keywords and phrases might be missing (e.g., “adverse childhood experiences”), they are still represented in the results due to semantic overlap and collocation with each of the keyword sets. The unsupervised classification of publications was further validated by examining the distribution of flagged policing and life-course publications over 30 topic clusters identified in the corpus (Figure S1).

[5] Note that local searches will not necessarily identify the path with the largest aggregate SPLC.

[6] While topic-irrelevant edges are deactivated for main path analysis, they are reactivated once the main paths of each (sub)field have been identified. This ensures all existing cross-citations are represented in the networks we use to examine communication between policing and life-course research.

[7] Nagin & Pogarsky (2001) cite a considerable amount of life-course literature, and are historically involved in life-course criminology, so it is likely that the language used in the abstract is evocative of life-course research without being explicitly related.