ISSN:

Recent advances in Evolution of Education and Outreach

Commentary       Open Access      Peer-Reviewed

The Developmental Similarity (Continuity) Construct: A Commentary

Peter V Paul*

Professor & Special Education Program Chair, Ohio State University, Department of Educational Studies, 305 Annie & John Glenn Avenue, PAES 354, Columbia

Author and article information

*Corresponding author: Peter V Paul, Professor & Special Education Program Chair, Ohio State University, Department of Educational Studies, 305 Annie & John Glenn Avenue, PAES 354, Columbia, E-mail: [email protected]
Received: 20 May, 2026 | Accepted: 09 June, 2026 | Published: 10 June, 2026
Keywords: Developmental similarity (continuity) hypothesis; Reading disability; Science of reading

Cite this as

Paul PV. Professor & Special Education Program Chair, Ohio State University, Department of Educational Studies, 305 Annie & John Glenn Avenue, PAES 354, Columbia. Recent Adv Evol Educ Outreach. 2026;3(1):25-30. Available from: 10.17352/raeeo.000010

Copyright License

© 2026 Paul PV. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Abstract

The primary focus of the present article is to describe the contributions of the developmental similarity (continuity) hypothesis to understanding the acquisition of beginning and mature reading comprehension skills. Some principles are common across several reading similarity models such as, for example, Scarborough’s Reading Rope Model, Simple View of Reading, and Paul’s Qualitative Similarity Hypothesis. In general, it is argued that, with qualifications, the developmental similarity (continuity) hypothesis applies to all readers, including struggling readers and those with disabilities. Buttressed by the Science of Reading, these models assert that all individuals need to acquire the same fundamentals for learning to read, to subsequently read to learn. These models present critical implications for instruction, curriculum, and assessment. Also discussed are challenges to the models resulting in a nuanced consensus, mainly with respect to variations in language access, self-regulation, and sociocultural variables. In essence, the developmental similarity construct has contributed to a shift from explanations of reading acquisition, based predominantly on deficits or disabilities, to an emphasis on interventions that incorporate relevant instructional strategies, based on fundamentals. Recommendations for further exploration are presented in the concluding section.

The developmental similarity (continuity) hypothesis: a commentary

One of the most important debates of the 20th century and continuing into the 21st is whether the development of students with disabilities and those who are struggling (poor) readers is similar to or different from that of students who are typical language/literacy learners. This debate, along with legal and political forces, provided the impetus for the inclusion education movement in the United States [1]. The outcomes of this ongoing dialogue have critical implications for research and practice, particularly for the focus of this article—understanding the acquisition of reading skills.

The purpose of this professional commentary is to discuss the background and major tenets of the developmental similarity (continuity) hypothesis and relate this construct to extant theories of reading acquisition. After presenting a few challenges, such as those associated with sociocultural variables, it is argued that these challenges do not alter the major tenets of the similarity (continuity) hypothesis for the acquisition of reading skills. However, these challenges do present some evidence that developmental similarity models are not uniformly applicable across all readers without qualifications. The article ends with recommendations for further research and practice.

Method

The selection of articles for this manuscript is based on a methodology labeled a professional review [2,3]. Professional reviews are typically attempts to provide and clarify an overall conceptual picture for a broad topic such as language or literacy development. Authors select and synthesize a representative sample of publications to support their arguments.

As is the case for any type of literature review, some limitations should be mentioned. One limitation is that such a review can be biased because it is based on the selection and interpretation of research by the author. Other authors may render different interpretations and question the selection of sources. Another limitation of the present review is that there was no attempt to evaluate the quality indicators/technical merits of the selected works. Nevertheless, it is argued that the present review should facilitate and guide future explorations, necessary for improving the reading skills of children who are struggling readers or who have formally diagnosed disabilities.

Background of the developmental similarity (continuity) hypothesis

For much of the 20th century, there was a prevailing assumption that the developmental trajectory of children with disabilities was radically different from that of typical learners without disabilities [1,4]. This assumption seemed to support, in part, the development and implementation of “separate curricula” and specialized, individualized instructional methods for these children. In addition, the assumption tended to justify their placement in separate education programs such as special schools and institutions or self-contained classrooms in public schools.

Cracks in the assumption began to appear before the passage of PL 94-142 in 1975 and its subsequent amendments [1]. There was evidence that children with disabilities can benefit from being in classrooms with peers without disabilities [5]. Additional support for the inclusion of these children in general education classrooms was provided by scholars who argued that their acquisition of knowledge is developmentally similar to that of typical children [4]. In short, these children should have access to and will benefit substantially by being exposed to the same general education curricula as typical children without disabilities.

Strong early support for the similarity hypothesis can be seen in the work of Lenneberg [6,7] for language development and in the work of Stanovich for reading development [8-10]. There are several other scholars who have provided evidence for this hypothesis, for example, Adams, Goswami [11,12], and Perfetti [13]. The works of Lenneberg and Stanovich, discussed in the subsequent sections, are selected here as exemplars. Specifically, they provided the impetus, in my view, for the development of reading similarity models and have influenced my own Qualitative Similarity Hypothesis for d/Deaf and hard-of-hearing children and adolescents [14,15,].

The work of lenneberg

[6] examined the language acquisition processes of children with intellectual/cognitive disabilities. Interestingly, Lenneberg attempted to clarify the biological foundations of language acquisition with his participants. He focused on examining the critical period hypothesis (i.e., optimal period for language development) and argued that language was driven by biological maturation processes rather than solely by general intelligence.

This led to an ingenious method of matching groups of children on mental age, rather than chronological age. There were (and still are) concerns with the determination of mental age via the use of intelligence tests or even Piagetian tasks [16]. Nevertheless, Lenneberg was able to demonstrate that the language acquisition of children with intellectual/cognitive disability was developmentally/qualitatively similar (i.e., manner, process), albeit quantitatively delayed (i.e., slower rate, amount) with respect to mental age. He also asserted that children with cognitive impairments can acquire complex grammar and other language variables with adequate language input during the critical period (i.e., typically up to puberty). These findings have been corroborated by more recent research, including for children who are struggling language and literacy learners as well as those with intellectual/cognitive disabilities [17,18].

The work of stanovich

Stanovich’s developmental lag hypothesis focused on the early acquisition of reading skills before the third grade (in the USA) [8,9]. This hypothesis was an attempt to provide explanations for reading difficulties and to ascertain developmental trajectories in struggling readers, particularly those with “garden-variety” reading disabilities (i.e., non-dyslexic). In essence, Stanovich proposed a critical or optimal period for the development of beginning reading skills—essentially up to about the third-grade level. Because reading materials become exponentially more difficult and complex after the third grade, Stanovich argued that children need to learn to read by the third-grade level (i.e., reach a level of competence). Otherwise, they will not be able to read to learn, and they will continue to fall behind. This hypothesis provided a broader framework for Stanovich’s Matthew Effects—which states that good readers continue to develop higher levels of literacy whereas poor or struggling readers fall further behind grade level, resembling a spiraling downward trend.

Using reading-level matched designs (i.e., matched on reading ability), Stanovich was able to compare older struggling readers to younger skilled readers. Focusing on the development of phonology or phonologically related access skills, it was demonstrated that good readers possess adequate word identification skills, which facilitate their development of adequate reading comprehension skills. However, Stanovich also cautioned that reading comprehension requires more than adequate word identification skills.

In essence, Stanovich’s work provided support for developmental similarity (continuity) models. He reasoned that poor older readers are developmentally “younger” readers. They follow the same developmental trajectory, but they are progressing at a slower quantitative rate. This distinguishes reading similarity models from the so-called deficit models, which assert that reading differences may also be qualitative, not just quantitative, and that poor readers do not adhere to a similar pathway compared to that of good readers.

With intensive early intervention programs, it may be possible for poor or struggling readers to “catch up”. In general, the evidence for Stanovich’s model is mixed; however, it provides the strongest evidence for problems with decoding (access skills) and even vocabulary development as it affects reading. Recent research on the model suggests that there is some evidence for specific deficits, suggesting qualitative differences, especially for some dyslexic readers [19]. In essence, it is also important to consider sociocultural factors such as the quality of instruction and socioeconomic factors. As discussed later, sociocultural factors present unique challenges for developmental similarity hypotheses. These factors, specifically those associated with neurocognitive research findings, have qualified the similarity construct, leading to the proposal of a developmental similarity (continuity) construct.

Reading models

The similarity (continuity) construct

Most reading models, influenced by the National Reading Panel and the Science of Reading construct, implicitly endorse a developmental similarity (continuity) framework [20,21]. The similarity (continuity) reading models selected as exemplars in this section are the Simple View of Reading [22], Scarborough’s Rope Model (2001), and the Qualitative Similarity Hypothesis [15]. These selected models attempt to explain the acquisition of literacy for both typical and struggling readers by demonstrating that there are similar underlying cognitive processes, such as decoding and comprehension. This acquisition process is qualitatively or developmentally similar (i.e., similar stages, effective strategies), albeit there may be quantitative (i.e., rate or amount) differences due to the efficiency of the use of skills and the influence of sociocultural factors (e.g., language competency, home environment).

Based on extensive research, the National Reading Panel (NRP; 2000) delineated the major components for the development of early reading skills. These components include phonemic awareness, phonics, fluency, vocabulary, and text comprehension. With the work of the NRP as a springboard, the Science of Reading (https://www.thereadingleague.org/wp-content/uploads/2026/05/science-of-reading-defining-guide-ebook.pdf) represents an expansion of reading research, which not only confirmed the original NRP findings but also discussed applications to older and diverse learners, including those with disabilities such as dyslexia, intellectual/cognitive disabilities, autism, and deafness.

The Science of Reading inspired more nuanced models of reading development, exemplified by constructs such as the Simple View of Reading [22], Scarborough’s Rope Model (2001), and Paul’s Qualitative Similarity Hypothesis [15]. These post-NRP models demonstrated that explicit and systematic instruction of the fundamentals (i.e., the NRP components) provides benefits for several struggling readers. In addition, a case was made that the same underlying reading processes (e.g., language and cognitive) apply to most, if not all, learners. This implies that learners follow a similar developmental trajectory in the acquisition of reading skills.

Both the Simple View of Reading [22] and Scarborough’s Rope Model (2001) emphasize two broad processes—decoding and language comprehension. However, Scarborough’s model views reading as an integration of decoding processes involving phonology and sight recognition and language comprehension processes such as vocabulary, syntax, reasoning, and background knowledge. Thus, there are subcomponents within each major component: decoding and language comprehension. As readers develop, these processes become “braided” or integrated. Scarborough’s model is more nuanced than the SVR. Neither the SVR nor Scarborough’s model stipulates separate processes for children with disabilities, nor do they propose alternative pathways. Both models can be used to argue implicitly for the similarity construct.

Related to the SVR and the Reading Rope Model is Paul’s Qualitative Similarity Hypothesis (QSH; [15]. This framework asserts that the development of the English language and literacy skills of d/Deaf and hard-of-hearing (d/Dhh) individuals is qualitatively similar to that of typical (hearing) learners. Despite the fact that the development might be quantitatively slower or delayed, d/Dhh readers proceed through similar stages, employ similar successful strategies, and make similar errors as younger literacy learners. The QSH supports the use of evidence-based practices that are effective for typical learners, but these practices need to be differentiated with respect to the individual profiles of d/Dhh learners. It has been argued that d/Dhh learners can progress and perhaps “catch up” if there are early intervention programs with intensive educational supports [23,24].

The QSH also aligns with the Science of Reading with its emphasis on intensive instruction of the same foundational skills explicated for typical literacy learners. Similar to “hearing” learners, d/Dhh students need to acquire the fundamental structures of English, including phonological processes, and other skills (e.g., vocabulary, comprehension) to learn to read and, specifically, to reach a mature reading competency stage [25]. In essence, the same reading “science” applies to d/Dhh children and adolescents, focusing on English language and reading-related components. There is no strong support for a different, exclusive sign-based methodology [24,26-28]. Although, at present, there have not been any scholarly discussions, it is suggested that the QSH can serve as a basis for implementing Scarborough’s Reading Rope with d/Dhh students

Challenges to the similarity models

Recent studies and meta-analyses strongly support a developmental similarity (continuity) hypothesis; that is, there is a continuity or generalization of extant reading theory that can be applied across populations regardless of ethnicity, gender, sexual orientation, or disability. This is in contrast to previous assumptions of fundamentally different reading mechanisms. However, there are qualifications, especially from neuroscience studies, related to sociocultural variables such as language background, home and cultural values, socioeconomic status, and school literacy practices [17,29,30,]. Although this recent line of research supports a broadly shared underlying developmental system, some scholars still assert that readers may demonstrate qualitatively distinct cognitive profiles or pathways due to sociocultural variables ( [31,32].

The sociocultural variables seem to qualify a strict developmental similarity conclusion. For example, Paul’s QSH asserts that d/Dhh learners, like “hearing” learners, work with similarities with respect to structures (grammar), patterns (letter-sound relations), and reasoning (use of prior knowledge and metacognition). According to the QSH (and other similarity models), these language and cognitive fundamentals do not change across contexts. Nevertheless, sociocultural variables impact how d/Dhh students access, develop, and interpret the fundamentals. In fact, given d/Dhh students’ challenges in developing a first language and other issues such as impoverished home environments and inadequate teacher-student relations, they may have difficulties in acquiring the fundamentals, even with supports. Similar to other struggling readers [30], some d/Dhh students may exhibit different neural patterns—albeit this needs to be confirmed by further investigations.

In essence, sociocultural and other related variables do not alter the fundamentals of learning to read, which are the same for all readers of, for example, English. These fundamentals include [33].

Knowledge of the language of print (i.e., components of english)

Metalinguistic awareness of language- and print-related factors (e.g., of letters, sounds, letter-sound relationships [phonology, orthography], functions of print [pragmatics], words [semantics], sentences [syntax]

Phonological-based working memory or phonological memory (and processes)

Comprehension capabilities (i.e., the development of textual, intertextual, and cultural prior knowledge, metacognitive, and self-regulatory skills)

However, it needs to be acknowledged that sociocultural variables have qualified the similarity hypothesis. Neurocognitive studies have indicated that not all reading differences are purely quantitative. Struggling readers often demonstrate less efficient processing and slower neural connectivity, and this seems to be due to an alteration with respect to, at least, phonological processes [29-37].

This leads to the proffering of a “qualified continuity model,” indicating that the development of reading skills is continuous, but requires specific attention to individual learners’ profiles. These profiles may reveal specific qualitative differences based on the impacts of sociocultural variables. Nevertheless, a nuanced developmental similarity model should attempt to integrate the notions of continuity and constrained differences. To ensure the mature development of reading, researchers and educators need to address the “differences.” The integration of “continuity” and “difference” should provide a better understanding of the reading process, rather than the proffering of exclusive explanations.

This manuscript examined the viability of the developmental similarity (continuity) hypothesis to better understand the reading acquisition process of children and adolescents. Foundational support for the hypothesis was presented via the seminal works of Lenneberg and Stanovich. It was noted that extant reading frameworks averred that common underlying language and cognitive processes apply broadly, with qualifications, across all literacy learners, including individuals with diagnosed disabilities such as intellectual/cognitive, autism, and deafness.

In sum, there is strong support for the developmental similarity models; however, attention needs to be focused on the individual profiles of readers. There may be qualitative differences due to sociocultural and neurocognitive factors such as language background and neural processing abilities. These factors influence the manner in which readers access and develop the foundational reading skills and can impact the subsequent acquisition of mature literacy skills. Nevertheless, the general core requisites remain consistent, indicating that the effectiveness of the models requires an integration of universal shared processes and individual variability. This integration should result in the improvement of reading outcomes, including for students with disabilities.

Recommendations for further research are as follows:

  • There is a need for longitudinal and multidisciplinary research, using complex research designs based on extant reading models, to obtain a deeper understanding of the developmental pathways of diverse learners.
  • It is critical to conduct neurocognitive investigations to explore the interactions of the physiological processes of the brain with the sociocultural elements of the learning and home environments. This might shed more light on the nature of any qualitative differences and how these differences can be addressed.
  • Teacher education programs should provide a curriculum that leads to a broader understanding of the development of literacy skills. This curriculum should include information on extant reading models from multidisciplinary sources.

There continues to be a decline in reading achievement scores, even after the pandemic, for both typical (https://nces.ed.gov/nationsreportcard/reading/) and struggling readers, including those with disabilities (https://www.k12dive.com/news/NAEP-special-education-scores-decline-assessment/739715/). Theorists and scholars should focus on implementation in educational settings, thereby reducing the gap between research and practice. Critical reading skills are becoming even more important for addressing complex problems of technologically infused societies.

References

  1. Winzer MA. From integration to inclusion: A history of special education in the 20th century. Washington (DC): Gallaudet University Press; 2009. Available from: https://www.researchgate.net/publication/287247838_From_integration_to_inclusion_A_history_of_special_education_in_the_20th_Century
  2. Gall MD, Gall JP, Borg WR. Educational research: An introduction. 8th ed. Boston (MA): Allyn & Bacon; 2007. Available from: https://www.scirp.org/reference/referencespapers?referenceid=1035831
  3. Gall MD, Gall JP, Borg WR. Applying educational research: How to read, do, and use research to solve problems of practice. 7th ed. Boston (MA): Pearson/Allyn & Bacon; 2015. Available from: https://www.amazon.in/Applying-Educational-Research-Problems-Practice/dp/0133831574
  4. Stainback W, Stainback S. A rationale for the merger of special and regular education. Except Child. 1984;51(2):102-111. Available from: https://doi.org/10.1177/001440298405100201
  5. Dunn LM. Special education for the mildly retarded—is much of it justifiable? Except Child. 1968;35(1):5-22. Available from https://doi.org/10.1177/001440296803500101
  6. Lenneberg EH. Biological foundations of language. New York (NY): Wiley; 1967.
  7. Rymer R. Annals of science: A silent childhood–I. New Yorker. 1992 Apr 13:41-81. Available from: https://www.newyorker.com/magazine/1992/04/13/i-a-silent-childhood
  8. Stanovich KE. Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Read Res Q. 1986;21(4):360-407. Available from: https://doi.org/10.1598/RRQ.21.4.1
  9. Stanovich KE. Children’s reading and the development of phonological awareness. Detroit (MI): Wayne State University; 1988.
  10. Stanovich KE. Progress in understanding reading: Scientific foundations and new frontiers. New York (NY): Guilford Press; 2000. Available from: https://www.routledge.com/Progress-in-Understanding-Reading-Scientific-Foundations-and-New-Frontiers/Stanovich/p/book/9781572305656
  11. Adams MJ. Beginning to read: Thinking and learning about print. Cambridge (MA): MIT Press; 1990. Available from: https://cpin.us/sites/default/files/fcab_resources/fcab_res_langlit/fcab_ll_bg/Beginning_To_Read.pdf
  12. Goswami U. Phonology, reading development, and dyslexia: A cross-linguistic perspective. Ann Dyslexia. 2002;52(1):141-163. Available from: https://doi.org/10.1007/s11881-002-0010-0
  13. Perfetti CA. Reading ability. Oxford (UK): Oxford University Press; 1985.
  14. Paul PV. The qualitative similarity hypothesis: A commentary. Hum Res Rehabil. 2021;11(2):56-61. Available from: https://doi.org/10.21554/hrr.092101
  15. Paul PV, Wang Y, Williams C. Deaf students and the qualitative similarity hypothesis: Understanding language and literacy development. Washington (DC): Gallaudet University Press; 2013. Available from: https://www.researchgate.net/publication/289432719_Deaf_students_and_the_qualitative_similarity_hypothesis_Understanding_language_and_literacy_development
  16. Russo N, Kaplan-Kahn EA, Wilson J, Criss A, Burack JA. Choices, challenges, and constraints: A pragmatic examination of the limits of mental age matching in empirical research. Dev Psychopathol. 2021;33(2):727-738. Available from: https://doi.org/10.1017/S0954579420001480
  17. Nilsson K, Elwer A, Messer D, Danielsson H. Cognitive and language abilities associated with reading in intellectual disability: A systematic review and meta-analysis. Remed Spec Educ. 2025:1-14. Available from: https://doi.org/10.1177/07419325251328644
  18. Spencer M, Wagner RK. The comprehension problems of children with poor reading comprehension despite adequate decoding: A meta-analysis. Rev Educ Res. 2018;88(3):366-400. Available from: https://doi.org/10.3102/0034654317749187
  19. Pfost M, Hattie J, Dörrer T, Artelt C. Individual differences in reading development: A review of 25 years of empirical research on Matthew effects in reading. Rev Educ Res. 2014;84(2):203-244. Available from: https://doi.org/10.3102/0034654313509492
  20. Moje EB, Afflerbach PP, Enciso P, Lesaux NK, editors. Handbook of reading research. Vol 5. New York (NY): Routledge; 2020. Available from:  https://www.routledge.com/Handbook-of-Reading-Research-Volume-V/Moje-Afflerbach-Enciso-Lesaux/p/book/9781138937376
  21. Snowling MJ, Hulme C, Nation K, editors. The science of reading: A handbook. Oxford (UK): Wiley-Blackwell; 2022. Available from:  https://content.e-bookshelf.de/media/reading/L-579231-b064811ccb.pdf
  22. Hoover WA, Gough PB. The simple view of reading. Read Writ. 1990;2(2):127-160. Available from: https://doi.org/10.1007/BF00401799 
  23. Aslan-Bağcı O. Is the qualitative similarity hypothesis applicable to Turkish d/Deaf and hard of hearing students? Am Ann Deaf. 2025;170(3):141-166. Available from: https://doi.org/10.1353/aad.2025.a985471
  24. Paul PV. An adequate model for the development of English reading skills. Hum Res Rehabil. 2024;14(2):284-292.
  25. Hayes P, Arnold P. Are hearing-impaired children’s reading delayed or different? J Res Read. 1992;15(2):104-116. Available from: https://doi.org/10.1111/j.1467-9817.1992.tb00026.x
  26. Mayer C, Trezek BJ. Is reading different for deaf individuals? Reexamining the role of phonology. Am Ann Deaf. 2014;159(4):359-371. Available from: https://doi.org/10.1353/aad.2014.0032
  27. Paul PV, Yan P. The effects of American Sign Language on English reading proficiency. Am Ann Deaf. 2023;166(5):745-760. Available from: https://doi.org/10.1353/aad.2023.0010
  28. Trezek BJ, Mayer C. Reading and deafness: State of the evidence and implications for research and practice. Educ Sci. 2019;9(3):216. Available from: https://doi.org/10.3390/educsci9030216
  29. Frei N, Willinger D, Haller P, et al. Toward a mechanistic understanding of reading difficulties: Deviant audiovisual learning dynamics and network connectivity in children with poor reading skills. J Neurosci. 2025;45(17). Available from: https://doi.org/10.1523/JNEUROSCI.1119-24.2025
  30. Tan LH, Perfetti CA, Ziegler JC, McCandliss BD. Editorial: Neural bases of reading acquisition and reading disability. Front Neurosci. 2023;17:1147156.  Available from: https://doi.org/10.3389/fnins.2023.1147156
  31. Dimitra S. Developmental dyslexia: Convergence of all theories towards a multifactorial model, and the role of ICTs. GSC Adv Res Rev. 2023;17(1):47-60. Available from: https://doi.org/10.30574/gscarr.2023.17.1.0386
  32. Lorusso ML, Toraldo A. Revisiting multifactor models of dyslexia: Do they fit empirical data, and what are their implications for intervention? Brain Sci. 2023;13(2):328. Available from: https://doi.org/10.3390/brainsci13020328
  33. Paul PV. Qualitative similarity hypothesis. In: Paul PV, Moores DF, editors. Deaf epistemologies. Washington (DC): Gallaudet University Press; 2012. p. 179-198.
  34. National Reading Panel. Report of the National Reading Panel: Teaching children to read—An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Jessup (MD): National Institute for Literacy; 2000.
  35. Scarborough HS. Connecting early language and literacy to later reading (dis)abilities: Evidence, theory, and practice. In: Neuman S, Dickinson D, editors. Handbook for research in early literacy. New York (NY): Guilford Press; 2001. p. 97-110. Available from: https://www.researchgate.net/publication/284286826_Connecting_early_language_and_literacy_to_later_reading_disabilities_Evidence_theory_and_practice
  36. Stainback W, Stainback S. A rationale for the merger of special and regular education. Except Child. 1984;51(2):102-111. Available from: https://doi.org/10.1177/001440298405100201
  37. Stanovich KE, Nathan R, Zolman J. The developmental lag hypothesis in reading: Longitudinal and matched reading-level comparisons. Child Dev. 1988;59(1):71-86. Available from: https://doi.org/10.2307/1130390
 

Help ?