Introduction
In his 1982 presidential address, Donald E. Walker provided a critical historical narrative of the Association for Computational Linguistics (ACL), focusing on the pivotal year of 1968. This period marked the transition from the Association for Machine Translation and Computational Linguistics (AMTCL) to the ACL, a change that signaled a fundamental shift in the field's identity. Walker’s reflection chronicles how the community moved from a narrow focus on translation applications toward a broader, interdisciplinary science of language and information.
The ALPAC Report and the Identity Crisis
The mid-1960s were a tumultuous time for computational linguistics. The field was initially centered almost exclusively on Machine Translation (MT). However, the publication of the 1966 ALPAC (Automatic Language Processing Advisory Committee) report created a significant crisis. The report argued that MT had failed to deliver on its early promises, leading to drastic funding cuts.
Walker notes a paradox regarding the ALPAC report: while it is often remembered for killing MT funding, its authors actually intended to advocate for increased support for basic research. Unfortunately, the report’s skepticism "frightened rather than challenged" funding agencies, forcing the community to confront a deep tension between immediate practical applications and fundamental scientific inquiry.
From Translation to Computational Linguistics
The 1968 name change from AMTCL to ACL was far more than a rebranding exercise. It represented a strategic pivot. The community recognized that to solve complex applications like translation, it first needed to address "basic issues" in linguistics and information retrieval. This shift allowed the field to establish a more stable, interdisciplinary foundation that was not solely dependent on the success of a single application.
Walker describes the "proper procedures" of constitutional amendments required for this change, illustrating the deliberate institutional effort to redefine the scope of the society. This evolution expanded the field's remit to include stylistic analysis, information retrieval, and what we now recognize as applied natural language processing.
Building an Interdisciplinary Hub
A key theme in Walker’s address is the vision of the ACL as a "central point" connecting diverse fields, including engineering, the humanities, social sciences, and education. Under Walker’s leadership, the ACL sought to move from being an "unofficial observer" to a constituent society within major international computing and AI organizations, such as the American Federation of Information Processing Societies (AFIPS) and the International Joint Conferences on Artificial Intelligence (IJCAI).
This "global perspective" was intended to embed computational linguistics within the broader scientific community. Interestingly, Walker injects a moment of levity into this history, attributing the slow adoption of his "global message" in 1968 to a "most horrible meal" served at an ACL banquet, which supposedly left members too demoralized to rally to the cause.
The Human Element and the Application Gap
Walker argued that the future of the field lay in studying how people organize and use information. He suggested that observing human interaction with computational systems would revolutionize the understanding of the field more than pure theory alone.
Despite the progress made since 1968, Walker acknowledged a persistent "application gap" between basic research and practical usage. He concluded by urging the community to reach out to practitioners who were performing computational linguistics work without identifying with the field, thereby grounding the science in real-world application.
Conclusion
"A Society in Transition" serves as a reminder that the foundations of the ACL were built on a necessary pivot from narrow application to broad scientific inquiry. By navigating the political and scientific challenges of the 1960s, the organization established the interdisciplinary framework that continues to define computational linguistics today.
Summary updated