National Instructional Materials Accessibility Standard Report—Version 1.0

Click here for this report in Microsoft Word format

Important Note: Part B of this report, the NIMAS Technical Specification, is no longer current.  Please refer to the current NIMAS 1.1 specification for more information.

Panel Recommendations

The Office of Special Education Programs at the U. S. Department of Education funded the National Center on Accessing the General Curriculum at CAST (Center for Applied Special Technology) to convene a panel of experts, the National File Format Technical Panel, and to establish technical specifications for a voluntary National Instructional Materials Accessibility Standard (NIMAS). NIMAS will facilitate the provision of accessible, alternate format versions of print textbooks to PreK–12 students with disabilities. The application of NIMAS will provide standardized source files from which accurate and reliable alternate formats may be developed and distributed to qualified students with disabilities.

The forty-member Technical Panel, representing educators, publishers, technology specialists, and advocacy groups, began its work in November, 2002 and developed recommendations during three, public face-to-face meetings in January, March, and June, 2003, supplemented by extensive teleconference and online discussions. The Panel defined a set of Governing Principles detailing the needs of students with disabilities with respect to obtaining access to print instructional materials, and guided by these principles achieved consensus on the following recommendations.

• To eliminate the need for repetitious and ineffective transformations (print-to-Braille; print-to-DTB, etc.) the agreed upon file format must be sufficiently flexible to create multiple student-ready transformations (Braille, Digital Talking Book, etc.) from the same source file.
  • The Panel recommends that an application of the ANSI/NISO Z39.86 standard, a subset of the DTBook Element Set and its associated Package File, be identified as NIMAS version 1.0. Textbook publishers will provide this baseline element set, and third-party conversion entities will supplement it with optional elements. Digital source files containing both elements will then be transformed into student ready versions. Guidelines for the application of both baseline and optional elements are included in the report. Please refer to Section III, Technical Specifications Recommended by the Panel for more detail.
  • In order to facilitate the creation of text equivalents, textbook publishers will also provide PDF files containing embedded images.
• In order to ensure the quality, consistency, and timely distribution of accessible curriculum materials, the NIMAS must address the inefficient policies and procedures hampering the current development and distribution systems. The panel recommends:
  • A two-year phase in of the NIMAS specification.
  • Conformity of all aspects of NIMAS-compliant content transformation to existing ANSI/NISO Z39.86 specifications and recommendations.
  • And creation of a NIMAS Committee to achieve:
    • The development of additional technical documentation and models that ensure the valid and consistent application of both baseline and optional elements.
    • The continued alignment of the NIMAS standard with ANSI/NISO Z39.86 and subsequent versioning of the NIMAS standard.
    • Capacity-building activities across a broad range of disability advocacy organizations to further acquaint them with the benefits of accessible alternate versions of textbooks.
    • A research agenda to investigate the pedagogical efficacy of the increased availability of accessible learning materials.
• To eliminate the current struggle to implement solutions on a student-by-student basis (particularly for blind and print impaired students), the NIMAS must establish a unified approach for the development of accessible learning materials. And it must facilitate, at the state level, the development of a systemic approach to providing accommodations for diverse learners by increasing their capacity to acquire and transform digital curricular materials.
  • The Panel acknowledges that the distribution of student-ready accessible alternate versions of print textbooks must respect existing copyright regulations (Section 121, the Chafee Amendment).
  • The Panel recommends active consideration of alternate content distribution models (Voluntary, Mandated, Free Market) that will effectively extend the availability of accessible alternate versions of textbooks to all students with disabilities and simultaneously ensure compensation to intellectual property holders and content developers.
  • The Panel unanimously recommends the national adoption of the NIMAS version 1.0 standard as the first step in ensuring the provision of appropriate and accessible textbooks to all students.

The recommendations of the Technical Panel reflect the sustained and active collaboration that has distinguished this panel's work. All panel members, represented organizations, providers of public commentary and the Office of Special Education Programs should be commended for helping to achieve this precedent setting consensus. The panel unanimously believes that the adoption of the National Instructional Materials Accessibility Standard Version 1.0 will significantly enhance the opportunity of students with disabilities to access, participate and progress in the general education curriculum.

National File Format Technical Panel: Contact List

U.S. Department of Education: Ex-Officio

• Robert Pasternack, Assistant Secretary, Office of Special Education and Rehabilitative Services
• Stephanie Lee, Director, Office of Special Education Programs
• Louis Danielson, Director, Office of Special Education Programs, Research-to-Practice Division.

NCAC Project Officer

• Bonnie Jones, Education Program Specialist, Office of Special Education Programs

CAST/National Center on Accessing the General Curriculum (NCAC)

• Chuck Hitchcock, NCAC Project Director, CAST, Wakefield, MA
• David Rose, NCAC Principal Investigator, CAST, Wakefield, MA
• Bob Dolan, Project Technical Support, CAST, Wakefield, MA

Consumer/Advocacy Group

• Mary Ann Siller, Director, National Education Program, American Foundation for the Blind, Dallax, TX
• Paul Schroeder, Vice President of Government Relations, American Foundation for the Blind, Washington, D.C.
• John D. Kemp, President/CEO, HalfthePlanet Foundation, Washington, D.C.
• Curtis Chong, Director of Field Operations and Access Technology, Iowa Department for the Blind, Des Moines, IA
• Mark Richert, Executive Director, Association for Education & Rehabilitation of the Blind & Visually Impaired, Alexandria VA
• Melanie Brunson, Director of Advocacy and Governmental Affairs, American Council of the Blind, Washington, D.C.
• Barbara Cheadle, President, National Organization of Parents of Blind Children, Division of the NFB, Baltimore, MD
• Robin Church, Assistant Vice President for Education Learning Disabilities Association, Executive Director of Kennedy Krieger School, Kennedy Krieger Institute, Baltimore, MD
• Randall Boone, Council for Learning Disabilities, University of Nevada, Las Vegas, Department of Education, Las Vegas, NV
• Sheldon Horowitz, Director of Professional Services, National Center for Learning Disabilities, New York, NY
• Barbara Raimondo, J.D., Director of Public Affairs, American Society for Deaf Children, Washington Grove, MD
• Alice Parker, Assistant Superintendent and Director Special Education, National Association of State Directors of Special Education, California Department of Education, Sacramento, CA
• Susan LaVenture, Executive Director, National Association for Parents of Children with Visual Impairments (NAPVI), Watertown, MA
• Gus Estrella, Policy Analyst, United Cerebral Palsy Association (UCPA), Washington, D.C.
• Muffi Lavigne, Information and Referral Coordinator, United Cerebral Palsy Association (UCPA), Washington, D.C.
• Michael M. Behrman, Professor, Graduate School of Education, George Mason University, Fairfax, VA
• Michael Wehmeyer, Associate Professor, Special Education, University of Kansas, Lawrence, KS
• Irvin Shapell, Publisher, Woodbine House, Bethesda, MD
• Jim Allan, Texas School for the Blind and Visually Impaired, Austin, TX

Technical Group

• Joe Sullivan, President, Duxbury Systems, Inc., Westford, MA
• Larry Skutchan, Technology Project Leader, The American Printing House for the Blind, Inc., Louisville, KY
• Eileen Curran, Vice President, Education Services, National Braille Press, Boston, MA
• Bill Raeder, President, National Braille Press, Boston, MA
• James Pritchett, Project Manager, Digital Audio Recording for the Blind and Dyslexic, Princeton, NJ
• Steve McBride, Chair, Electronic Instructional Materials Committee, National Association of State Textbook Administrators, East Charleston, WV
• Charles E. Mayo, Assistant Division Director/ Distribution/Accessible Textbooks, Texas Education Agency, Textbook Administration Division, Austin, TX
• Kim Hartsell, Project Director, Georgia Project for Assistive Technology, Forest Park, GA
• Greg Pisocky, Business Development Manager Adobe Systems, Inc., Vienna, VA
• Gary Moulton, Microsoft Accessible Technology Group, Microsoft Corporation, Redmond, WA
• Madeleine Rothberg, Director of Research and Development, National Center for Accessible Media, WGBH Boston, Boston, MA
• Michael Moodie, Research and Development Officer, National Library Service for the Blind and Physically Handicapped, Washington, D.C.
• Rick Ferrie, Vice President/General Manager, Mazer Corporation, Boston, MA
• Gene Golovchinsky, Sr. Research Scientist, FX Palo Alto Laboratory, Inc., Palo Alto, CA
• Dave A. Schleppenbach, ghBraille, LLC, West Lafayette, IN
• Martin Hensel, President, Texterity Inc., Southborough, MA

Feasibility Group

• Skip Stahl, Co-Director, Universal Learning Center, CAST, Wakefield, MA
• John Roberts, National Institute of Standards and Technology, Gaithersburg, MD
• Margaret E. Bausch Department of Special Education and Rehabilitation Counseling, University of Kentucky, Lexington, KY
• Steve Driesler, Executive Director, School Division, Association of American Publishers, Washington, D.C.
• Pearce McNulty, Houghton Mifflin Company, Boston, MA
• Alex Mlawsky, Glencoe/McGraw Hill, Columbus, OH
• George Kerscher, Senior Officer, Accessible Information DAISY Consortium, Open eBook Forum, Recording for the Blind and Dyslexic, Missoula, MT
• Martha Minow, Harvard Children's Initiative, Harvard Law School, Cambridge, MA

Table of Contents

1. Background Information
   1. Relationship to NCLB
   2. The Need for a National Standard
   3. The Goals
   4. The Process
   5. The Technical Panel
   6. Summary of Meetings
      1. Inaugural Meeting
      2. Meeting 2
      3. Meeting 3
2. Technical Panel Consensus and Understanding
   1. Identifying Student Needs
   2. Baseline Format for the National Instructional Materials Accessibility Standard
   3. Alignment with Existing Standards
   4. Limitations of and Restrictions on the National Instructional Materials Accessibility Standard
3. Technical Specifications Recommended by the Panel
   1. Governing Principles for the National Instructional Materials Accessibility Standard
      1. Principle 1: The NIMAS should be sufficiently extensible to support transformations into multiple presentation formats (e.g., text to properly formatted braille, text equivalents of non-text elements, etc.).
      2. Principle 2: The NIMAS should support the inclusion of all media that are contained in the printed work (e.g., text, graphics, charts, tables, etc.).
      3. Principle 3: The NIMAS should provide for the incorporation of structural information about the print work (e.g., chapter, section, and subsections; paragraphs; page breaks and page numbers; line numbers where essential; indexes; tables of content, etc.).
      4. Principle 4: The NIMAS format should facilitate the efficient and automated conversion of print material into more accessible formats.
      5. Principle 5: The NIMAS should accommodate all fields of study, including those that require special notation (e.g., math and music) as standards become available.
   2. Technical Specifications
      1. Application of ANSI/NISO Z39.86 Standard for the National Instructional Materials Accessibility Standard
      2. The Baseline Element Set
      3. The Optional Elements and Guidelines for Use
      4. Package File
4. Ensuring the Success of the National Instructional Materials Accessibility Standard
   1. Quality Assurance
   2. Documentation and Training
   3. Distribution
   4. Workflow Under Existing Copyright Restrictions
   5. File Validation and Delivery
   6. Maintenance of the NIMAS Specification
5. Additional Recommendations
   1. Timeframe for Implementation
   2. Improving Awareness
   3. All Students with Disabilities
      1. Copyright Constraints
      2. Free Market Distribution of Accessible Versions
      3. Moving Beyond Accessibility
   4. Research on Efficacy
      1. Accessibility: Suggested Questions for future inquiry
      2. Availability: Suggested Questions for future inquiry
6. Resources
   1. Appendix A: Technical Panel Detailed Meeting Summaries:
      1. Meeting 1: January 7, 2003
      2. Meeting 2: March 11, 2003
      3. Meeting 3: June 9 & 10, 2003
   2. Appendix B
      1. Usage Guidelines
   3. Appendix C
      1. Optional Elements
   4. Appendix D
      1. Glossary of Acronyms and Terms

Scope of Work: The Technical Panel has been charged with providing the Secretary of Education with a set of technical specifications to facilitate the efficient delivery of accessible instructional materials, a timeline for the implementation of the proposed standards, and process for assessing the success of standards implementation.

I. Background Information

A. Relationship to NCLB

The No Child Left Behind Act of 2001 (NCLB) established strict new requirements for schools. These requirements affected teacher qualifications, instructional programs and strategies, student testing and test reporting, and support for students not meeting standards and students with special needs. But most fundamentally NCLB increased schools' accountability for all students' educational progress. Such accountability was first legislated with the landmark 1997 Amendments to the Individuals with Disabilities Education Act (IDEA), which demanded access, participation, and progress for students with disabilities in the general education curriculum, as well as participation in large-scale assessments. NCLB cemented and even intensified this outlined responsibility of schools to provide students with disabilities with the same quality education as their peers.

But there is still an achievement gap separating students with disabilities and their peers (Thurlow, Wiley & Bielinsky, 2003). Existing general education curricula and the accountability systems that accompany them are not usually designed to achieve or measure results for children with disabilities. Indeed, students with disabilities are rarely included during any phase of the development, adoption, or validation of curricula. Thus, most general education curricula lack research-based alternative methods, materials, and assessments that support the progress of diverse learners.

One of the most significant shortcomings of general education curricula is their widespread use of rigid, print-based materials. Following the passage of IDEA, it has become essential that all students have access to the general curriculum, and thus to the print materials of which it is largely composed. But for many students with disabilities, the limitations of print technology raise barriers to access, and therefore to learning. For students who cannot see the words or images on a page, cannot hold a book or turn its pages, cannot decode the text or comprehend its syntax, or cannot engage with words on paper, printed text is a serious impediment. These students may each experience different challenges and require different supports to extract meaning from this medium, but the barrier they face is one and the same—printed text. In a print-based curriculum these students may be misidentified as unable to learn when in reality they simply do not have appropriate learning tools and materials to access learning.

Teachers must make their best efforts to diminish the barriers raised by print textbooks through adaptations, accommodations, and the development of alternative systems and programs. These locally developed alternatives are costly and inefficient, lack a research base and systematic development, and create yet another separate track rather than true engagement in the general curriculum. In order to resolve these issues and realistically implement the priorities of NCLB and IDEA, students with disabilities need access to more flexible versions of materials, and they need this access at the same time that print versions are available to their non-disabled peers.

Modern digital materials can present the same content as printed books but in a medium that is much more flexible and accessible. For students who cannot see the words or images, the digital version can more easily be converted to braille or voice with text descriptions of the images. For students who cannot hold the printed book or turn its pages, the virtual pages of a digital book can be turned with a slight press of a switch. For students who cannot decode the text, any word can be automatically read aloud. For students who lack the background vocabulary in the text, definitions (in English or another language) can be provided with a simple click. Moving beyond accessibility, digital texts can also be embedded with supports for syntax, semantics, and comprehension (Boone & Higgins, 1993; Dalton, Pisha, Eagleton, Coyne, & Deysher, 2001 MacArthur & Haynes, 1995).

The advantage of digital versions is that these alternatives, and many others, can be available on an individual basis—available for students who need them, invisible or non-distracting for those who don't. They enable teachers to individualize materials in previously unimaginable ways (Hay, 1997; Lewin, 2000; MacArthur & Haynes, 1995). Such customizable alternatives can substantially reduce the barriers found in traditional texts. Mounting research evidence demonstrates the benefits of using such digital materials in the classroom (Barker & Torgesen, 1995; Bottge, 1999; Dalton et al., 2001; Erdner, Guy, & Bush, 1998; MacArthur & Haynes, 1995; Wise, Ring, & Olson, 1999).

B. The Need for a National Standard

As a result of the new accountability for all students, the availability of accessible core curriculum materials has become critically important to state and local educational agencies. Authentic access, participation and progress of students with disabilities in the general education curriculum will not occur until accessible, flexible, and customizable learning resources are widely available to all students, particularly those with print disabilities.

For several reasons, few students with disabilities presently have access to the highly functional accessible books they need. In some cases the problem is knowledge-related—many educators do not understand the issue of access or the potential solutions that are available. For the majority of students however, lack of access is rooted primarily in the absence of a uniform development and distribution system for accessible versions of print textbooks, an absence that precludes their timely delivery.

The existing development and distribution systems rely on inefficient policies and procedures that raise barriers rather than opportunities. These barriers hinder success at every level of the system.

Publishers generally do not manufacture and distribute fully accessible digital versions of print materials for use by all students with disabilities. Although laws in a number of states require publishers to provide digital versions of their instructional materials to third parties (authorized entities as defined by the Chafee Amendment copyright exemption) for conversion into braille and other accessible formats for students with print disabilities, publishers face inefficiencies based on current regulatory and technological requirements. Unclear administrative processes and the lack of a universally-accepted file format at the state level often hamper publishers' compliance efforts. Without a universal file format it is difficult for publishers to comply (different states, disability organizations, teachers and students all request different versions and formats). Additionally, the current system promotes redundancy in efforts to convert the digital versions into accessible formats, increases the costs to the states for acquiring these materials, and causes unnecessary delays in receiving the accessible formats.

Moreover, these laws do not address the needs of all children with disabilities. The Chafee Amendment, enacted in 1996, is narrowly tailored to serve a small subset of students with disabilities (i.e., those with print disabilities). Qualification for eligibility for alternate format materials references 36 CFR 701.10 Title 36 Code of Federal Regulations and identifies:

• Blind persons whose visual acuity, as determined by competent authority, is 20/200 or less in the better eye with correcting glasses; or whose widest diameter of visual field subtends an angular distance no greater than 20 degrees.
• Persons whose visual disability, with correction and regardless of optical measurement, is certified by competent authority as preventing the reading of standard printed material.
• Persons certified by competent authority as unable to read or unable to use standard printed material as a result of physical limitations.
• Persons certified by competent authority as having a reading disability resulting from organic dysfunction and of sufficient severity to prevent their reading printed material in a normal manner.

Pursuant to the Chafee Amendment, the current distribution structure for accessible formats intentionally bypasses the marketplace and provides the materials through a third-party authorized entity without compensation to the copyright holder. The needs of a broader set of students with disabilities could be met through a variety of free market solutions. For example, publishers could directly manufacture and distribute fully accessible versions of their materials to serve the needs of this population. Alternatively, a licensing system could be created to simplify the process for third parties to obtain permission from, and provide compensation to, the textbook copyright holders in order to make and distribute accessible versions of the print textbooks.

Current publishing practices also diminish the financial incentives to make and distribute fully accessible materials, because it is not yet commonplace for publishers to acquire the digital rights to the materials they produce, or the images that they purchase to include in textbooks. These materials are often used in multiple editions of books. Consequently, it is difficult for publishers to acquire the digital rights of pre-existing material after the original licensing arrangement has been set. Thus, creating a marketplace, driving demand, and providing publishers with the financial incentives to produce and distribute fully accessible digital materials in a manner generally consistent with copyright law are critical to ensuring full access to print materials for all children with disabilities.

Not-for-profit agencies and their designated vendors, referred to as authorized entities in the Chafee Amendment, have emerged to develop and distribute more accessible versions of published materials on behalf of individuals with disabilities. The largest of these, like Recording for the Blind and Dyslexic (RFB&D) and American Printing House for The Blind (APH), are national and have long served students by creating alternate formats of print materials. Under the Chafee Amendment, these formats are restricted to, braille, audio, or digital text. This limitation means that these authorized entities are currently not able to expand their production and distribution of accessible materials to serve a broader range of students. Many of these agencies have begun migrating to digital versions as a better way to serve their clients, but the lack of a standard source format from publishers for these versions has hampered their progress. Because different publishers tend to use different proprietary formats or templates in their digital production and distribution, it is difficult for authorized entities to develop an efficient process for transforming materials. Most find it easier to begin with the printed book itself rather than a digital version supplied by the publisher, opting to recreate a digital version from the printed book.

Beyond the inefficiencies created by the lack of a consistent and uniform publisher source format, additional redundancy of effort exists since most authorized entities do not, at present, have in place any mechanism for or incentive to share marked up digital files. The implementation of a uniform publisher source format will provide the technological foundation for subsequent sharing of the digital source file among authorized entities, an approach that would further decrease the current inefficiencies associated with the creation of accessible versions of print textbooks.

Outside of the not-for-profit arena, for-profit conversion companies like Texterity, Data Conversion Laboratory, Mazer Corporation, ghBraille LLC, Duxbury Systems and others would also benefit from the adoption of a uniform publisher source format in the work they do as a part of commercial product development or as subcontractors in the conversion cycle.

Because schools and school districts generally do not get accessible versions of their curricular materials directly from the same sources as they get their regular materials, they must work with authorized agencies to obtain accessible materials—or else create them themselves. Either path is complicated and time consuming. Districts must identify the format or formats that both meet the needs of individual students and work with the technologies available in their various schools and classrooms. They must select a vendor or process for creating the necessary formats. After ordering materials they must develop a repository and distribution system to match versions to individual students and technologies. Finally, they need to develop local capacity for supporting teachers in using these accessible versions. The abundance of technologies and formats greatly complicates this already drawn out process and delays the delivery of materials until a point at which all too often the need has passed.

Teachers have limited choices for acquiring accessible versions for their classrooms. If teachers work in districts or within states that have a system of guidelines in place to define the process of identifying what books are needed and how orders are processed and alternative formats delivered to students, they can receive audio or braille versions of textbooks. If teachers work in districts without an established procedure, they can identify a format that is appropriate for their student(s) and compatible with their existing classroom technology and then find a vendor or repository that can supply it. A second option for such teachers is to work with an in-district authorized entity, a qualifying special educator, for example, to create, in a time consuming and often error-ridden process, (e.g., without access to a licensed braille transcriber), a digital version of the print textbook themselves. Neither of these present options had proven to be successful in providing high-quality accessible textbooks to students in a timely manner. Moreover, teachers are not able to obtain materials for every student who needs them in an accessible format, due to the copyright and market constraints referenced previously.

While there are many barriers to accessibility, the problems that are caused by multiple file formats are particularly frustrating. But they are also easily remedied by adoption of a standard file format. The adoption of a common or standard format is a simplifying step that has been crucial to progress in many other fields from railroads (adopting a common track gauge) to video technology (adopting a common format for DVD and HDTV). Similarly, defining a common National Instructional Materials Accessibility Standard (NIMAS) for the provision of digital source versions of print textbooks would greatly stimulate progress in accessibility. Existing and emerging standards related to electronic publishing support the designation of a flexible core format that separates the content of instructional materials from its presentation, allowing for output of multiple accessible versions, including braille, large print and audio, from a single source. With that single change, a number of barriers at many points in the educational system can be addressed.

With one clear and consistent file format to produce, publishers would be able to deliver a high quality digital version of their print instructional materials expeditiously and simultaneously to all authorized entities for further conversion and distribution.

Authorized entities would be able to efficiently transform these materials into an accessible version (i.e., accessible digital versions and printed braille) and expeditiously deliver them to schools and school districts.

It should be noted, however, that the NIMAS initiative is, at present, indelibly linked to existing copyright law and its exemption, the Chafee Amendment. This linkage will continue to require a multi-step process in order to ensure that students are provided access to accessible versions of print textbooks. The multi-step process involves the delivery of source files (NIMAS version 1.0-compliant) to third parties (authorized entities) for conversion markup. Once this secondary (optional) markup has been completed, the source files are ready for transformation into student ready formats such as braille, Digital Talking Book and other specialized formats. For a more detailed discussion of this process, please refer to Section IV A of this report.

With the provision of materials in one common digital format to third party conversion organizations, the likelihood that teachers could consistently get accessible materials in a timely fashion and in a format that could be readily implemented in the classroom increases significantly. Students would finally get the accessible materials they need when they need them.

Several key states, California, Texas, New York, and Kentucky, have existing or have enacted new or amended legislation requiring the consideration of accessibility in the textbook procurement process. The intent of these state legislative initiatives is to ensure that students with disabilities have access to appropriate learning resources at the same time that traditional materials are made available to their non-disabled peers. But the existence of a national standard for the provision of digital source files for print textbooks, from which accessible student versions can subsequently be created, is vital to this objective. In fact, it is the only way to achieve the U.S. Department of Education's goal of an efficient system for production and distribution of accessible core curriculum instructional materials. A NIMAS has the potential to significantly enhance the learning opportunities available to all students and is therefore of significant importance to publishers, technology specialists, teachers, advocacy groups, and students.

C. The Goals

The goals of this initiative were several-fold. The overarching goal was to make appropriate and accessible versions of print textbooks available to every student who needs them. This goal is currently undermined by the inefficiency of the development and distribution system for these materials, which leaves schools struggling to provide instructional materials on a student-by-student basis (particularly for students with print disabilities). Because the system's inefficiency derives in large part from the multiplicity of file formats, the specific aim of this work was to make progress toward standardization.

The general approach was to select a panel of stakeholders that would develop and detail a proposed NIMAS that would limit the use of diverse file formats in favor of a single source format sufficiently flexible to create multiple output transformations (braille, Digital Talking Book, etc). This one step would drive the creation of solutions to problems existent throughout the accessible curricular materials development and distribution system, achieving the following objectives:

• Improving the efficiency of third party relationships by reducing the need of publishers to produce different formats for different instances.
• Making the process of transformation and delivery by authorized entities more efficient by reducing the need for repetitious and often ineffective transformations (print-to-braille; print-to-ebook, etc.) and establishing a unified approach for the development of accessible learning materials.
• Simplifying the acquisition process for schools and teachers.
• Simplifying the implementation process for teachers.

Development of a NIMAS raises a number of challenging and important questions that must be addressed for such a standard to be worthwhile and effective. While pursuing its targeted aim of developing a NIMAS, the Technical Panel sought to specifically address the following questions and problems:

• What structure will best ensure that states, schools, and students have access to accessible digital curricular materials in a timely fashion? Student needs for digital curricular materials change over time. The crux of the problem today is an inefficient system that cannot regularly support "just in time" delivery. As a result, by the time students receive their materials they often are no longer appropriate.
• How do we overcome the constraint imposed by the Chafee Amendment copyright exemption, which limits the distribution of accessible content to a subset of the special needs population? Although the Chafee Amendment was originally crafted as an approach to ensure the practical availability of accessible versions to students who need them, its definition of qualifying students does not cover many in need.
• How do we reconcile the technological conflict between accessibility and Digital Rights Management (DRM), technology-based security that limits unauthorized distribution in order to facilitate the exercise of marketing rights by the copyright owner? At the present time, for DRM to be effective, the text in ebooks must be locked to prevent copying. Popular assistive technology tools such as screen readers and supported reading software are incompatible with locked text.
• How useful are accessible instructional materials in increasing the achievement of students with disabilities? This and other research questions related to the consideration of students with disabilities in standards-based instruction—standards, materials, instructional practice and assessment—needs to be addressed.

Ultimately addressing these important questions and achieving the specific aim of developing a NIMAS will facilitate, at the state level, the development of a systemic approach to providing accommodations for diverse learners, ensuring the quality, consistency, and timely distribution of accessible curriculum materials to students.

D. The Process

At the direction of and in consultation with the U.S. Department of Education, the National Center for Accessing the General Curriculum (NCAC), in collaboration with the National Institute of Standards and Technology (NIST) at the Department of Commerce, assembled a Technical Panel consisting of 40 members representing consumers, technical experts, and feasibility experts.

Identified since its inception as the National File Format Technical Panel, the goal of this group has also been referred to as the National File Format. Throughout this report the designation National File Format and its acronym (NFF) have been replaced with the designation National Instructional Materials Accessibility Standard or NIMAS to align with the formal language in the charge from the U S. Department of Education.

The Panel was convened at a series of three meetings (all open to the public, their dates, times, and locations posted on the NCAC Web site: http://nimas.cast.org) over a 6-month period. Each of the three Panel meetings was focused on the contributions of members representing each of the Panel's subgroups (consumers, technical experts and feasibility specialists). The first meeting concentrated on identifying the needs of students with disabilities with respect to accessible materials and emphasized the contributions of the consumer representatives. The second meeting concentrated on identifying the technical issues involved in establishing a NIMAS, and emphasized the contributions of the technical representatives. The third meeting concentrated on presenting the feasibility issues (intellectual property concerns, market exigencies, etc.) and emphasized the contributions of the feasibility representatives. Detailed summaries of all three Panel meetings are available at the Web site referenced above and are attached to this report as Appendix A.

Throughout these meetings, the Panel developed and refined a proposed instructional materials accessibility standard in compliance with the constraints of 2U.S.C. 15a, "An Act to Provide Books for the Adult Blind" and a subsequent copyright exemption built upon that Act (e.g., Section 121, the 1996 "Chafee Amendment"), which authorize the provision of accessible versions of non-dramatic works without compensation to the copyright holder to individuals with qualifying print disabilities.

Before each meeting, group representatives were asked to submit specific information that would inform the discussion. After each meeting CAST consolidated a record of the discussion and circulated a draft of these items for review, discussion, and further revision. These subsequent communications occurred in an online forum and via follow-up telephone conferences.

E. The Technical Panel

In collaboration with NIST at the Department of Commerce, NCAC convened a forty member Technical Panel from each of the following groups:

• Organizations that represent consumers (e.g. advocacy groups for individuals with visual, print and other disabilities and state and local educational representatives).
• Organizations that represent parents of students with disabilities (e.g., blind, print and other disabilities).
• Instructional materials publishers or their organizational designees (e.g., School Division, Association of American Publishers).
• Assistive technology developers or their organizational designees (e.g., Assistive Technology Industry Association).
• Producers of materials in specialized formats (e.g., RFB&D, APH, etc.).
• State special education program administrators or their organizational designees (e.g., National Association of State Directors of Special Education).
• Special educators and local administrators.
• State textbook administrators from adoption states or their organizational designees (e.g., National Association of State Textbook Administrators) and equivalent personnel from non-adoption states.
• Developers of accessibility, publishing software and supporting technologies.
• Information technology standards organizations (e.g., The Access Board).
• Data transformation organizations with substantial experience in accessible file format preparation and automated conversion technologies.

Nominees to the Technical Panel were selected for their expertise in identifying and detailing the specific instructional needs of K–12 students with disabilities, their investment in increasing the availability of accessible digital materials, their experience in curriculum design, and their technical expertise.

In addition to delegates from the constituencies listed above, the Panel solicited commentary and guidance from NCAC partner members, including Harvard University Law School, and the Secretary of Education designated 3 employees from the Department as ex-officio members of the Technical Panel.

The charge of this Technical Panel was to present the Secretary of Education with "a set of technical specifications to facilitate the efficient delivery of accessible instructional materials, a timeline for the implementation of the proposed standards, and process for assessing the success of standards implementation."

F. Summary of Meetings

1. Inaugural Meeting

During the first meeting, held January 7, 2003, the Technical Panel addressed the needs of the Consumer constituency and began developing a list of the features and functions that the instructional materials accessibility standard must address. To inform this discussion each participating organization representing the consumer constituency was asked to submit a wish list of the features and functionality that would benefit their constituent students. During this same period, CAST gathered comments from interested consumers who were not members of the Technical Panel. Based on the discussion and input from this meeting, CAST developed a working list of proposed instructional materials accessibility standard features and functions, which was then circulated among the Panel members for comment and feedback. Panel members had the opportunity to discuss and revise the proposed working list through online discussions at the Online NFF Community established by CAST. Recommended features and functions were then approved at a brief follow-up meeting/telephone conference in early February.

For a more detailed summary of the inaugural meeting please see Appendix A.

2. Meeting 2

At the second meeting on March 11, 2003, the Technical Panel addressed the interests of the Technical Group and began the process of identifying the technical specifications for an instructional materials accessibility standard. Prior to the meeting, representatives from the Technical Group were asked to review existing file formats, accessibility guidelines and emerging technologies in the context of the wish list and submit their recommendations and concerns. As with the first meeting, CAST coordinated the comments and feedback of Panel members and circulated the proposed technical standards following the meeting. Technical Panel members had the opportunity to discuss and revise these proposed technical standards at the NFF Technical Panel Web site and during a follow-up meeting/telephone conference held at the end of March.

For a more detailed summary of the second meeting please see Appendix A.

3. Meeting 3

At the third and final meeting on June 9 and 10, 2003, the Technical Panel assessed the market influences, copyright constraints, and protocols and procedures necessary to ensure the widespread adoption of voluntary standards. Feasibility Group representatives were asked to submit their comments on this topic to CAST before the meeting. Following the meeting CAST once again circulated a proposed set of recommendations to the Technical Panel members and provided the Technical Panel with an opportunity to discuss and revise the recommendations through the NFF Web site and a follow-up meeting/telephone conference.

For a more detailed summary of the third meeting please see Appendix A.

II. Technical Panel Consensus and Understanding

During the course of the three convenings, the Technical Panel built a consensus around the need for a NIMAS to improve access to educational materials for children with disabilities. Consensus was reached on four issues: 1) guiding principles for a NIMAS; 2) baseline format for the NIMAS; 3) application of the format for the NIMAS; and 4) limitations of and restrictions on the NIMAS.

A. Identifying Student Needs

As set forth in greater detail in Section III below, the Technical Panel developed a set of principles to guide the adoption and implementation of the NIMAS. These governing principles reflect the Technical Panel's assessment of the barriers that students with disabilities face in obtaining access to print instructional materials. They attempt to define the boundaries of what it means to provide access to educational materials for students with disabilities. Alone, these principles are neither prescriptive nor restrictive; rather, they are intended to help those following the NIMAS to interpret whether a particular application satisfies the goal of providing access to instructional materials.

B. Baseline Format for the National Instructional Materials Accessibility Standard

Rather than develop a new standard, the Technical Panel opted to develop guidelines for the use and implementation of an existing standard. The primary reason for this decision was the time constraint on the panel for completing its report and recommendations, which precluded development of a completely new standard. The Technical Panel considered several different standards, including ANSI/NISO Z39.86, DocBook, and simple ASCII files. After significant discussion, it concluded that an extensible format should be adopted, primarily to ensure that the format would be sufficiently flexible to address the needs of a variety of children with disabilities, to provide output in a variety of formats, to expand to address new needs as they are identified, and to easily evolve over time. Additionally, the Technical Panel recognized that several existing standards had been developed with the intent to address many of the same needs identified by this panel. Several of these standards had already gained traction in segments of the education community and therefore had conversion houses and publishers who were already technically capable of providing instructional materials in these standards.

One standard in particular, ANSI/NISO Z39.86, was formally submitted to the Technical Panel for consideration by panel members who served both on the NIMAS Technical Panel and on the Z39.86 Advisory Committee.^* These Panel members provided this background information on the Z39.86 standard:

One key factor behind the ever-growing enthusiasm for Z39.86 is the support for the XML source content identified in DTBook. Demonstrations at the NIST 2001 Electronic Publishing conference showed a source file encoded in DTBook, and from that file braille, a Microsoft Reader version, and a Digital Talking Book were generated. Thus, DTBook enables the generation of many formats from a single source. The support for Z39.86 ranges across a variety of areas:

• National organizations who provide content to persons with disabilities, such as NLS, RFB&D, APH, American Foundation for the Blind (through AFB Press), and BookShare have implemented or are planning to implement the standard.
• Developers of braille translation software are building their tools to accommodate Z39.86.
• The AFB Textbooks and Instructional Materials Solutions Forum is poised to support Z39.86 in the further development of their training program.
• Internationally, a DAISY XML Techniques working group is focused on implementation and training for the Z39.86 standards.
• Conversion houses that transform publishers' content have expressed great interest in DTBook and the Z39.86 standard.
• Recording software developers have plans to implement the standard.
• Multimedia developers are looking at identifying features that can enhance the standard.
• Playback and Reading System developers are planning to implement the standard.
• Open source developments are targeting Z39.86 for their activities.
• The Advisory Committee is developing supporting software and valid sample implementations.
• The United Nations has expressed great interest in the DAISY standards for making information accessible for all.

Based on these factors, the Technical Panel agreed to designate the NIMAS as an application of the DTBook element set of the ANSI/NISO Z39.86 standard. The identification of the NIMAS as an application of an existing standard allowed the Panel to separate the DTBook element set into two distinct categories:

1. The Baseline element set: digital files marked up with these tags are provided by publishers and comprise NIMAS version 1.0
2. The Optional element set: required in order to subsequently create an accessible representation of the textbook, the NIMAS-conformant digital files are marked up by authorized entities with these additional tags to make them ready for transformation into student ready versions (braille, Digital Talking Book, etc.)

The ANSI/NISO Z39.86 standard "defines the format and content of the electronic file set that comprises a digital talking book (DTB)" (http://www.niso.org/standards/resources/Z39-86-2002.html). Commonly referred to as "DAISY" (Digital Accessible Information System), the ANSI/NISO Z39.86 standard provides the technical protocol for creating accessible digital versions of print works. The following section is excerpted from the DAISY Structure Guidelines (available online at
http://www.daisy.org/publications/guidelines/sg-daisy3/structguide.htm#contents).

The DAISY DTB is a collection of digital files (from this point onward referred to simply as "files") that provides an accessible representation of the printed book for blind, visually-disabled, and print-disabled users. These files may contain digital audio recordings of human speech, marked-up text, and a range of machine-readable files.

The structure of the book is designated by the XML tags and is accessible to the reader by use of a browser or a playback device. The DAISY DTB utilizes the technology of the Internet with some specialized applications added to provide greatly improved access to the information.

DAISY 3 supports any of the following classes of DTB:

• Audio with Title element only: DTB without structure. This is the simplest class of DTB and is used for books where structure will not be applied. The XML textual content file may not be present, or if it is, contains only the title of the book, and other required notation. The book must be read linearly. Direct access to points within the DTB is not possible.
• Audio with NCX only (see "The NCX" below): DTB with structure. The XML textual content file, if present, contains only the structure of the book and may contain links to features such as narrated footnotes, etc. This is the most common form of DTB and is ideal for stand-alone players.
• Audio with NCX and partial text: DTB with structure and some additional text. The XML textual content file contains only the structure of the book and the text of components where keyword searching and direct access to the text would be beneficial (e.g., index, glossary, etc.).
• Audio and full text: DTB with structure and complete text and audio. This form of a DTB is the most complex but provides the greatest level of access. The XML textual content file contains the structure and the full text of the book. The audio and the text are synchronized.
• Full text and some audio: DTB with structure, complete text and limited audio. The XML textual content file contains the structure and the text of the book. The audio files contain recordings of parts of the text. This type of DTB could be used for a dictionary where only pronunciations were provided in audio form.
• Text and no audio: E-text with structure. The XML textual content file contains the structure and text of the book. There are no audio files.

XML provides the producer with the ability to structure a book in great detail. Compared to HTML markup, XML increases markup options and makes more detailed structure and proper nesting possible.

A DTB produced under DAISY 3 consists of some or all of the following files:

• A Package File (drawn from the Open eBook Forum™ (OEBF) Publication Structure 1.0.1), containing administrative information about the DTB, the files that comprise it, and how these files interrelate.
• A textual content file containing some or all of the text of the book with appropriate markup.
• Audio files containing the human voice recording of the book.
• SMIL (Synchronized Multimedia Integration Language) file(s) containing information linking the audio and textual content files.
• NCX, a file containing all points in the book to which the user may navigate.

In addition to the extensive audio supports for textual content (both digitized human voice and synthetic speech is supported), the ANSI/NISO Z39.86 (DAISY 3) standard incorporates accommodations for other media as well. Image files, pictures, charts, graphs, tables, maps, etc. can be included in DTBs along with their text equivalents (captions, long descriptions). ANSI/NISO Z39.86 incorporates Synchronized Multimedia Integration Language (SMIL), which allows for the addition of other accessible media into the Digital Talking DTB. For more information, see
http://www.niso.org/standards/resources/Z39-86-2002.html#SMIL.

Because the ANSI/NISO Z39.86 standard is so complex, the Technical Panel determined that the complete standard exceeds the capabilities of publishers at this time. Consequently, the broader standard will be treated as afterwork, easily integrated at a later time by the appropriate entity.

C. Alignment with Existing Standards

Identifying the NIMAS as an application of the existing ANSI/NISO Z39.86 standard creates alignment with all existing national and international guidelines and standards in the areas of digital formats and accessibility. As referenced by its identifying acronym, ANSI/NISO Z39.86 has been formally recognized as:

• An American National Standard Developed by the National Information Standards Organization (NISO).
• Approved March 6, 2002 by the American National Standards Institute (ANSI).

The development of ANSI/NISO Z39.86 by the DAISY Consortium has consistently built on existing standards work. Following the adoption of SMIL as a standard in 1998 by the World Wide Web Consortium (W3C), this specification, along with HTML, was subsequently incorporated into the DAISY 2.0 standard, wholly aligning the DAISY effort with the W3C. XML support (also a W3C standard) was added to the DAISY specification as version 2.01 in 1999. In 2001 DAISY version 2.02 was published, based on the W3C's XHTML 1.0 and SMIL 1.0 specifications. The ANSI/NISO Z39.86 2002 standard represents version 3.0 of the DAISY specification.

The Open eBook Forum (OeBF) was created in 1998 by publishing companies interested in the development of an open standard for electronic books, and promoted by NIST. In 1999, with the publication of version 1.0 of the eBook Publication Structure, the Open eBook Forum established accessible XML as a standard, thereby promoting interoperability between the work of the DAISY Consortium and the OeBF.

The designation of version 1.0 of the NIMAS as an application of ANSI/NISO Z39.86 2002 ensures that textbook files conforming to this specification will be structured in a manner that supports all existing standards and specifications relating to accessible digital materials.

D. Limitations of and Restrictions on the National Instructional Materials Accessibility Standard

Although the proposed NIMAS is a tremendous step forward, there are several important limitations and restrictions to recognize at this time.

• It is important to note that files produced in accordance with NIMAS, as developed by the Technical Panel, are not themselves usable by students with disabilities. In order to be made usable by students, these files must be converted into a format appropriate for braille output, an e-book reader, or some other alternate format. The purpose of the NIMAS is to ease these subsequent translations by enabling partial automation of the process, which would significantly decrease the time necessary to output materials in accessible formats.
• The agreed upon guidelines for NIMAS do not mandate the full markup of text necessary even for direct braille output or an electronic or talking book. The DTBook element set includes many other tags that are critical to this conversion but were deemed complex and requisite of skilled human intervention for appropriate tagging. Aligning with existing copyright law, the Technical Panel agreed to leave this additional work to third party conversion agents ("authorized entities") who in the current system provide such human intervention, preparing files for output in alternate student ready formats. Nothing in this agreement, however, limits textbook publishers from completing the full markup if they have both the intellectual property permissions and the financial incentives to do so.
• Nevertheless, concerns remains about who will perform this markup, the amount of time it will take, and the potential for duplication of efforts. Under the current system, each conversion agent would separately complete the markup of materials. In some cases, the conversion agent would extend the baseline NIMAS markup to make it fully conform to DTBook. In other cases, conversion agents may only add additional, proprietary tags (word pronunciation indicators for synthetic speech, for example), to complete the markup process at that stage. An optimal scenario would involve full DTBook mark up by conversion agents and the centralized availability of these fully marked up files.
• Without the availability of fully marked up files competing conversion agents could not be expected to share files with proprietary mark up, with the end result being that a file fully marked up for use by students in one state might not be available for use by students in another state. Each state would still have to contract for and pay for completed mark up, which would perpetuate the problem of delays in delivery of accessible versions of instructional materials. These are issues that need to be addressed in the future. Without resolving this issue of perpetuated redundancy the goal of establishing a truly viable national standard is significantly diminished.
• The Technical Panel recognizes that NIMAS will not meet the accessibility needs of all students with disabilities. During the Technical Panel meetings, there was a tension between developing a standard that is capable of providing for the accessibility needs of all students with disabilities, and the realities of the Chafee Amendment and other provisions of copyright laws that limit provision of accessible materials, in the absence of permission from the copyright holder, to children with recognized print disabilities. The publisher members of the Technical Panel expressed concern about adopting a standard that included accessibility features for individuals outside the narrow legal definition of an individual with a print disability and that could potentially encourage violations of the copyright laws. Other members of the panel, however, wanted to ensure that the standard would address the needs of a more comprehensive group of children with disabilities notwithstanding copyright constraints.
• This dilemma affects educators as well, because in spite of the copyright exemption they have a legal obligation to provide accessible educational materials to all students with disabilities. There is no law that restricts educators from requesting and purchasing accessible versions of educational materials from publishers for use by all students, including students with disabilities. But limited market demand restricts the feasibility of this approach—few publishers have commercially available accessible instructional materials that educators may purchase. Moreover, there is not currently a standard for accessible instructional materials for use by all students with disabilities. Because it was clearly beyond the scope of the NIMAS charge by the U.S. Department of Education to resolve this dilemma, the Technical Panel did not pursue a NIMAS that would include accessibility features for individuals outside the narrow legal definition of an individual with a print disability.

Endnotes
^*George Kerscher, James Pritchett and Michael Moodie

III. Technical Specifications Recommended by the Panel

Subsequent to the three meetings, the Panel finalized a set of governing principles that provide the accessibility and pedagogical foundation for the technical specifications, and the technical specifications themselves. These are presented in the following sections.

A. Governing Principles for the National Instructional Materials Accessibility Standard

1. Principle 1: The NIMAS should be sufficiently extensible to support transformations into multiple presentation formats (e.g., text to properly formatted braille, text equivalents of non-text elements, etc.).

Rationale for Principle 1: Files created in NIMAS are intended for transformation into formats for use by students with disabilities. At the present time this use is limited to students who qualify as "blind or other persons with disabilities" according to Section 121 of the Copyright Act. However, it is anticipated that NIMAS will eventually serve as a basis for the provision of accessible instructional materials to all students eligible for services and/or accommodations under IDEA, Americans with Disabilities Act or Section 504 of the Rehabilitation Act. Therefore the NIMAS specification needs to be sufficiently flexible to support both present and future transformation needs.

• NIMAS should facilitate the transformation of one media type into another: text to audio, text to properly formatted braille, etc.
• NIMAS should support the inclusion of simultaneous and synchronous presentation of multiple content media: video with text captioning, video description, etc.
• NIMAS should support the inclusion of layered content elements (e.g., text, graphics, etc.) to provide adjustable degrees of content complexity.

2. Principle 2: The NIMAS should support the inclusion of all media that are contained in the printed work (e.g., text, graphics, charts, tables, etc.).

Rationale for Principle 2: The original print content may be rendered auditorily, graphically or tactilely—or in any combination of those media—depending upon the needs of qualifying students. Therefore ALL original print content should be available to be transformed and their purpose (semantics) in the instructional sequence should be clearly represented.

• The NIMAS should support the inclusion of all graphical material: photos, images, charts, flow charts, tables, etc.
• The NIMAS should support the inclusion of high-resolution graphic elements capable of being scaled.
• The NIMAS should support the inclusion of semantic information that identifies the contextual intent of non-text elements (e.g., are graphics decorative, navigational, informative, etc.).

3. Principle 3: The NIMAS should provide for the incorporation of structural information about the print work (e.g., chapter, section, and subsections; paragraphs; page breaks and page numbers; line numbers where essential; indexes; tables of content, etc.).

Rationale for Principle 3: It is assumed that digital texts will be used side-by-side with print textbooks in the classroom, necessitating common points of reference. Currently, page breaks and paragraphs serve that function. The structural integrity of this reference system needs to be maintained to enable an accurate transformation into alternate media formats.

• NIMAS should support the preservation of meaning associated with layout if present in the print work.
• NIMAS should support flexible navigation among text and non-text elements to ensure that the student using the accessible version has an experience equitable to the student using the print work.
• NIMAS should support the inclusion of semantic information relevant to learning (summary, topic sentence, etc.).

4. Principle 4: The NIMAS format should facilitate the efficient and automated conversion of print material into more accessible formats.

Rationale for Principle 4: NIMAS should be capable of being rendered into multiple presentation formats (properly formatted braille, Digital Talking Book, etc.) in as automated a process as possible. To achieve this, the NIMAS should be based on a standardized structure that includes consistent rules, supports fine detail (granularity), and provides for future modifications.

• NIMAS should support successful machine validation for DTD conformance and adherence to NIMAS structure guidelines.

5. Principle 5: The NIMAS should accommodate all fields of study, including those that require special notation (e.g., math and music) as standards become available.

Rationale for Principle 5: Widely accepted structured markup already exists for some content areas—notably language arts, social science, history, subjects where the presentation of information is primarily text based with augmenting graphics. Standardized structured markup for mathematics, musical notation and science (alphanumeric character structures: equations, algorithms, formulas, etc.) should be incorporated into NIMAS when technically feasible.

B. Technical Specifications

1. Application of ANSI/NISO Z39.86 Standard for the National Instructional Materials Accessibility Standard

Because the DTBook is so comprehensive, the group opted to treat NIMAS as an implementation of the DTBook standard. It is not intended to narrow the scope of the DTBook standard, but to define the minimum actions that publishers must take in order to be in compliance with the standard. Additionally, the referenced guidelines are intended to establish best practices for how the tags should be applied to instructional materials.

The Baseline Element Set details the minimum requirement that must be delivered to fulfill the NIMAS standard. It is the responsibility of the publishers to provide this NIMAS-compliant XML file along with a separate PDF file with embedded images for reference in creating descriptions, and a NIMAS-compliant package file.

Content files meeting the NIMAS must be valid to the dtbook110.dtd [see ANSI/NISO Z39.86, Appendix 1
(http://www.niso.org/standards/resources/Z39-86-2002.html#DTBookDTD)]. In addition, files are required to use the tags from the Baseline Element Set when such tags are appropriate. Publishers may, and are encouraged to, augment the required Baseline Element Set with tags from the Optional Element Set (see Appendix C) as applicable. For the purposes of NIMAS, appropriate usage of elements (both baseline and optional) is defined by the "Guidelines for Use" (Appendix B), derived from the DAISY Structure Guidelines. Files that do not follow these guidelines in the selection and application of tags are not conformant to this standard.

2. The Baseline Element Set
a. Document-level tags (required to be valid XML)

Element	Description
dtbook	The root element in the Digital Talking Book DTD. <dtbook> contains metadata in <head> and the contents itself in <book>.
head	Contains metainformation about the book but no actual content of the book itself, which is placed in <book>. This information is consonant with the <head> information in xhtml, see [XHTML11STRICT]. Other miscellaneous elements can occur before and after the required <title>. By convention <title> should occur first.
book	Surrounds the actual content of the document, which is divided into <frontmatter>, <bodymatter>, and <rearmatter>. <head>, which contains metadata, precedes <book>.
meta	Indicates metadata about the book. It is an empty element that may appear repeatedly only in <head>.
title	Contains the title of the book but is used only as metainformation in <head>. Use <doctitle> within <book> for the actual book title, which will usually be the same.

b. Structure and Hierarchy

Element	Description
frontmatter	Usually contains <doctitle> and <docauthor>, as well as preliminary material that is often enclosed in appropriate <level> or <level1>. Content may include copyright notice, foreword, acknowledgments, table of contents, etc. <frontmatter>serves as a guide to the content and nature of a <book>.
bodymatter	Consists of the text proper of a book, as contrasted with preliminary material <frontmatter> or supplementary information in <rearmatter>.
rearmatter	Contains supplementary material such as appendices, glossaries, bibliographies, and indices. It follows the <bodymatter> of the book.
level1	The highest-level container of major divisions of a book. Used in <frontmatter>, <bodymatter>, and <rearmatter> to mark the largest divisions of the book (usually parts or chapters), inside which level2 subdivisions (often sections) may nest. The class attribute identifies the actual name (e.g., part, chapter) of the structure it marks. Contrast with <level>.
level2	Contains subdivisions that nest within <level1> divisions. The class attribute identifies the actual name (e.g., subpart, chapter, subsection) of the structure it marks.
level3	Contains sub-subdivisions that nest within <level2> subdivisions (e.g., sub-subsections within subsections). The class attribute identifies the actual name (e.g., section, subpart, subsubsection) of the subordinate structure it marks.
level4	Contains further subdivisions that nest within <level3> subdivisions. The class attribute identifies the actual name of the subordinate structure it marks.
level5	Contains further subdivisions that nest within <level4> subdivisions. The class attribute identifies the actual name of the subordinate structure it marks.
level6	Contains further subdivisions that nest within <level5> subdivisions. The class attribute identifies the actual name of the subordinate structure it marks.
h1	Contains the text of the heading for a <level1> structure.
h2	Contains the text of the heading for a <level2> structure.
h3	Contains the text of the heading for a <level3> structure.
h4	Contains the text of the heading for a <level4> structure.
h5	Contains the text of the heading for a <level5> structure.
h6	Contains the text of the heading for a <level6> structure.

c. Block Elements

Element	Description
author	Identifies the writer of a work other than this one. Contrast with <docauthor>, which identifies the author of this work. <author> typically occurs within <blockquote>.
blockquote	Indicates a block of quoted content that is set off from the surrounding text by paragraph breaks. Compare with <q>, which marks short, inline quotations.
list	Contains some form of list, ordered or unordered. The list may have intermixed heading <hd> (generally only one, possibly with <prodnote>) and an intermixture of list items <li> and <pagenum>. If bullets and outline enumerations are part of the print content, they are expected to prefix those list items in content, rather than be implicitly generated.
li	Marks each list item in a <list>. <li> content may be either inline or block and may include other nested lists. Alternatively it may contain a sequence of list item components, <lic>, that identify regularly occurring content, such as the heading and page number of each entry in a table of contents.
hd	Marks the text of a heading in a <list> or <sidebar>.
note	Marks a footnote, endnote, etc. Any local reference to <note id="yyy"> is by <noteref idref="#yyy">. [Attribute - id]
p	Contains a paragraph, which may contain subsidiary <list> or <dl>.
sidebar	Contains information supplementary to the main text and/or narrative flow and is often boxed and printed apart from the main text block on a page. It may have a heading <hd>.
cite	Marks a reference (or citation) to another document.
dd	Marks a definition of the preceding term <dt> within a definition list <dl>. A definition without a preceding <dt> has no semantic interpretation, but is visually presented aligned with other <dd>.
dl	Contains a definition list, usually consisting of pairs of terms <dt> and definitions <dd>. Any definition can contain another definition list.
dt	Marks a term in a definition list <dl> for which a definition <dd> follows.

d. Inline Elements

Element	Description
em	Indicates emphasis. Usually <em> is rendered in italics. Compare with <strong>.
q	Contains a short, inline quotation. Compare with <blockquote>, which marks a longer quotation set off from the surrounding text.
strong	Marks stronger emphasis than <em>. Visually <strong> is usually rendered bold.
sub	Indicates a subscript character (printed below a character's normal baseline). Can be used recursively and/or intermixed with <sup>.
sup	Marks a superscript character (printed above a character's normal baseline). Can be used recursively and/or intermixed with <sub>.
br	Marks a forced line break.
line	Marks a single logical line of text. Often used in conjunction with <linenum> in documents with numbered lines. [Include in baseline element set. Use only when line breaks must be preserved to capture meaning (e.g., poems, legal texts).]
linenum	Contains a line number, for example in legal text. [Include in baseline element set. Use only when <line> is used, and only for lines numbered in print book.]
pagenum	Contains one page number as it appears from the print document, usually inserted at the point within the file immediately preceding the first item of content on a new page. [NB: Only valid when includes id attribute].
noteref	Marks one or more characters that reference a footnote or endnote <note>. Contrast with <annoref>. <noteref> and <note> are independently skippable.

e. Tables

Element	Description
table	Contains cells of tabular data arranged in rows and columns. A <table> may have a <caption>. It may have descriptions of the columns in <col>s or groupings of several <col> in <colgroup>. A simple <table> may be made up of just rows <tr>. A long table crossing several pages of the print book should have separate <pagenum> values for each of the pages containing that <table> indicated on the page where it starts. Note the logical order of optional <thead>, optional <tfoot>, then one or more of either <tbody> or just rows <tr>. This order accommodates simple or large, complex tables. The <thead> and <tfoot> information usually helps identify content of the <tbody> rows. For a multiple-page print <table> the <thead> and <tfoot> are repeated on each page, but not redundantly tagged.
td	Indicates a table cell containing data.
tr	Marks one row of a <table> containing <th> or <td> cells.

f. Images

Element	Description
imggroup	Provides a container for one or more <img> and associated <caption>(s) and <prodnote>(s). A <prodnote> may contain a description of the image. The content model allows: 1) multiple <img> if they share a caption, with the ids of each <img> in the <caption imgref="id1 id2 ...">, 2) multiple <caption> if several captions refer to a single <img id="xxx"> where each caption has the same <caption imgref="xxx">, 3) multiple <prodnote> if different versions are needed for different media (e.g., large print, braille, or print). If several <prodnote> refer to a single <img id="xxx">, each prodnote has the same <prodnote imgref="xxx">.
caption	Describes a <table> or <img>. If used with <table> it must follow immediately after the <table> start tag. If used with <img> or <imggroup> it is not so constrained.

3. The Optional Elements and Guidelines for Use

Publishers are encouraged to apply markup beyond the baseline (required) elements. The complete DTBook Element Set reflects the tags necessary to create the six types of Digital Talking Books referenced in Section II and braille output. Because of the present necessity to subdivide the creation of alternate format materials into distinct phases, the Panel determined that baseline elements would be provided by publishers and optional elements would be added to the NIMAS-compliant files by third party conversion entities. In both circumstances the protocols for tagging the digital files should conform to the ANSI/NISO Z39.86 specification. For this reason the optional elements beyond the baseline set are included as an Appendix C, and content converters are directed to the DAISY Structure Guidelines
(http://www.daisy.org/publications/guidelines/sg-daisy3/structguide.htm) for guidance on their use.

4. Package File

A package file describes a publication. It identifies all other files in the publication and provides descriptive and access information about them. A publication must include a package file to be conformant to NIMAS. The package file is based on the Open eBook Publication Structure 1.2 package file specification (For most recent detail please see
http://www.openebook.org/oebps/oebps1.2/download/oeb12-xhtml.htm#sec2). A NIMAS package file must be an XML-valid OeB PS 1.2 package file instance and must meet the following additional standards:

The NIMAS Package File must include the following Dublin Core (dc:) metadata:

• dc:Title
• dc:Creator (if applicable)
• dc:Publisher
• dc:Date (Date of NIMAS-compliant file creation—yyyy-mm-dd)
• dc:Format (="NIMAS 1.0")
• dc:Identifier (a unique identifier for the NIMAS-compliant digital publication, e.g., print ISBN + "-NIMAS"—exact format to be determined)
• dc:Language (one instance, or multiple in the case of a foreign language textbook, etc.)
• dc:Rights (details to be determined)
• dc:Source (ISBN of print version of textbook)

And the following x-metadata items:

• nimas-SourceEdition (the edition of the print textbook)
• nimas-SourceDate (date of publication of the print textbook).

The following metadata was also proposed as a means of facilitating recordkeeping, storage and file retrieval:

• dc:Subject (Lang Arts, Soc Studies, etc.)
• nimas-grade (specific grade level of the print textbook, e.g.; Grade 6)
• nimas gradeRange (specific grade range of the print textbook, e.g.; Grades 4 - 5)

And additional suggestion references the use of:

• dc:audience:educationLevel (for the grade and gradeRange identifiers, noting that Dublin Core recommends using educationLevel with an appropriate controlled vocabulary for context, and recommends the U.S. Department of Education's Level of Education vocabulary online at http://www.ed.gov/admin/reference/index.jsp. Using educationLevel obviates the need for a separate field for gradeRange since dc elements can repeat more than once. A book used in more than one grade would therefore have two elements, one with value "Grade 4" and another with value "Grade 5."

A final determination of which these specific metadata elements to use needs to be clarified in practice.

The package manifest must list all provided files (text, images, etc.).

The package spine must reference all text content files in order. (Note: For purposes of continuity and to minimize errors in transformation and processing, the NIMAS-compliant digital text should be provided as a single document.)

IV. Ensuring the Success of the National Instructional Materials Accessibility Standard

A. Quality Assurance

As discussed at the final Technical Panel meeting, quality assurance can be divided into two distinct components: validation (ensuring that a file developed according to NIMAS specifications is a valid NIMAS file) and quality (ensuring that the elements -tags- applied in the validated file are done so according to recommended guidelines).

There is broad agreement that the validation process can be handled by an automatic (machine-based) procedure. NIMAS files can be parsed by an automatic validation engine that compares the submitted file (element existence, element hierarchy, proper use of attributes, etc.) to the required structure.

Establishing the quality of NIMAS files automatically is not felt to be currently feasible. Applying the elements and attributes with the accuracy necessary to maintain the integrity of the print work while simultaneously ensuring its accessibility is a challenging task. With respect to curriculum materials in general, and textbooks specifically, the structural components of the print work (headings, tables of contents, sidebars, etc.) provide a means of demarcating chapters, units, sections and their respective sub-components. In addition, some of these components and others (images, glossary terms, call out boxes, etc.) provide semantic richness, and the identification and placement of these elements must support their purpose within the context of the instructional design of the print work. The transformation and tagging of these components in the process of making them accessible must support the instructional intent to the greatest degree possible.

The best practice for assuring the quality of tagged files is for content developers to follow established guidelines. In this case, NIMAS is an application of the ANSI/NISO Z39.86 standard, and its associated elements are a subset of the DTBook Element Set. The required guidelines publication is the DAISY Structure Guidelines (http://www.daisy.org/publications/guidelines/sg-daisy3/structguide.htm). This document provides an element-by-element protocol for appropriate tagging, and content developers intending to submit NIMAS files should be familiar with its requirements. Beyond the delineation of each element and its application, however, content developers need to have an informed understanding of the rationale behind the recommended procedures. The DAISY Consortium has published a Statement of Principles for the Creation and Production of Accessible Books and Materials, reproduced below:

Version 1.0
Last Revised: June 14, 2002

This Statement of Principles represents the minimum standard to be met by Libraries for the Blind and producers of alternate format materials.

The DAISY Consortium endorse the following: Alternative format materials should be a faithful representation of a published work to ensure that those who are blind or print-disabled have access to the same information as the rest of society. Libraries or other producers of alternate format versions who adopt these Principles assure readers that these versions, regardless of their country of origin, represent the original published work as closely as possible and are accessible.

In terms of these principles, accessible is defined as follows. The print document is rendered in an alternative format such as braille, audio, large print or e-text in such a manner as to enable the print-handicapped reader to extract information from the document as easily and fully as possible.

DAISY supports the following principles in the creation and production of alternate format materials:

• Producers will respect and maintain the integrity of the published work to every extent possible. The producer will not alter the content or omit essential components unless absolutely necessary to provide access to the information.
• It is recognized that it may be necessary for producers to alter the intended order of the published work or add explanation of visual material so as to make the content accessible to the reader.
• If it is necessary to alter the presentation of the published work solely for the purpose of accessibility any information that a producer adds or omits must be identified as such.
• It is acknowledged that some producers will be required to produce only a portion of a published work. These partial books should be identified as such.
• It is recommended that producers of digital materials apply a structure to the published work to facilitate access and navigation by the reader.

Prepared for: DAISY Consortium and IFLA: Section of Libraries for the Blind Date: January 27, 1999 (Copyright © 2002 DAISY Consortium) http://www.daisy.org/publications/guidelines/sg-daisy3/principles.htm#title)

These principles provide a clear statement of the goals of creating accessible versions of print works, and content developers should actively apply these principles when creating accessible versions.

It should be noted that the term integrity in the DAISY statement of principles refers to the preservation of structural components and their order of presentation. With respect to preserving the instructional purpose of these structural components (and additional semantic elements), especially within the context of facilitating not just the access, but the achievement of students with disabilities, guiding principles have yet to be articulated.

Successful implementation of NIMAS will require the combination of expertise in the area of accessibility with expertise in the realm of instructional design. This is necessary to provide equitable curriculum resources that are not only usable by students with disabilities, but actually increase their achievement.

B. Documentation and Training

Because NIMAS is an application of an extant standard, considerable documentation, especially with respect to the tagging of text and structural elements already exists and is publicly available. Protocols for the use of elements, attributes and hierarchies need to be augmented with practical examples (models) of appropriate NIMAS structured markup. A number of complimentary and cooperative initiatives are currently underway to create these models and make them available to all stakeholders.

Samples of textbook transformations should be compiled across all grade ranges (PreK–12) and in as many subjects areas (language arts, social studies, etc.) as accepted mark up procedures exist. These models should be publicly available and open to ready analysis in order to encourage as widespread and uniform an application as possible.

Additional documentation needs to be developed pertaining to increasing the accessibility, placement and purpose of graphical elements in educational materials. The majority of prior accessibility initiatives aimed at utilizing the flexibility of digital media to increase accessibility have focused primarily on textual components. And so, guidelines, models and protocols for effectively rendering graphical elements in order to preserve their instructional purpose need to be developed.

C. Distribution

Determining the optimal workflow for implementing NIMAS on a national scale is outside the Panel's scope of work and therefore not part of the Panel's formal recommendation. However, all Technical Panel members felt that discussing distribution as a part of the final report was essential to facilitate the successful adoption of the standard by publishers, states, local education agencies and schools. Considerations reflect three categories of implementation:

• Mandated: If the NIMAS compliance mandate for states and publishers remains in the reauthorization of IDEA, the timely establishment of an efficient, secure and quality-assured workflow will be required.
• Voluntary: If the NIMAS specification emerges as a voluntary standard, publishers and states wishing to become early adopters will benefit from workflow recommendations made by the Panel.
• Free Market: If the NIMAS specification is to be the foundation element for source documents developed by curriculum publishers for subsequent transformation into accessible presentation formats available for direct sale to states, districts and schools, both publishers and education entities will benefit from workflow suggestions.

Within each of the three implementation categories delineated above, two possible storage and delivery options exist: centralized and distributed. For example, a centralized distribution system might exist as a delivery component of mandatory or voluntary implementation of the specification. A distributed system of delivery could also apply regardless of how the specification is implemented. Thus, the workflow suggestions that follow identify both types of storage and delivery options. Note: The following analysis of workflow processes was originally developed by Technical Panel members Dave Schleppenbach, Martin Hensel, and Rick Ferrie and submitted to the technical subgroup for consideration on April 14, 2003. Subsequent edits and additions have been made to reflect considerations that emerged following that submission.

D. Workflow Under Existing Copyright Restrictions

The workflow outlined below suggests a process for creating, validating and distributing NIMAS-compliant accessible textbooks within the confines of the existing Section 121 copyright exemption (Chafee Amendment). Within a Chafee-compliant distribution model the three steps referenced will be required regardless of whether the validation and delivery workflow is managed through a centralized system or a distributed system.

There are three major steps that must be taken in order for the student to get an actual accessible textbook:

• Publisher Markup (baseline NIMAS): this is the minimum requirement that must be delivered to fulfill the NIMAS standard. It is the responsibility of the publishers to provide this NIMAS-compliant XML file along with PDF files with embedded images for reference in creating descriptions. (Note: File nomenclature is yet to be determined.) A suggestion is the numbering of image files based on the page number that they appear in the book, followed by an ordering number for flow purposes (e.g., 13-1, 13-2, and so forth). All images should be included; making the proper resolution images and sorting through which images need to be conveyed in an alternative format is left up to the value-add stage. This core NIMAS file may be generated by curriculum publishers utilizing internal company resources or with base markup entities (third party organizations or companies hired by a publisher to create the NIMAS-compliant source files).
• Third-Party Markup (for specific implementations of NIMAS such as Braille, Digital Talking Book, Audio, etc.): this is the step needed for actual creation of the accessible file in the traditional modes of braille, Digital Talking Book, and Audio versions. This will most likely be done by the third party value-add organizations (authorized entities under the Chafee Amendment). However, there is nothing from a technical standpoint that would prevent publishers from taking on this task themselves (or in collaboration with a base markup or value-add entity), with the purpose of selling the accessible content along with the print content at cover price. Similarly, states with XML transformation capacity might choose to complete this stage of the process themselves.
• Delivery of Presentation Formats: the final student ready accessible version of the transformed textbook is delivered to state and local education agencies for monitored distribution to qualifying students following the requirements of Section 121 of the Copyright Act (Chafee Amendment). State or local education agencies, in collaboration with the value-add organizations, will be responsible for ensuring due diligence in maintaining compliance with the Chafee Amendment requirements.

It is important to re-emphasize the fact that both Steps 1 and 2 above are required to create the source files necessary for the subsequent creation of student ready alternative accessible textbooks. The recommended baseline markup is insufficient to create these accessible versions. While the establishment of a centralized repository (see below) would provide a mechanism for combining Steps 1 and 2 into a more cohesive whole, the segmentation of the process for creating accessible versions of textbooks will continue to persist until either a logical alternative or a clear market exigency evolves. In the meantime, this segmentation will continue to create inefficiency and redundancy.

E. File Validation and Delivery

Although formal consideration of the method of delivery for NIMAS compliant files is outside of the Panel's scope of work, discussion of how this process would occur was repeatedly raised during the three face-to-face Panel meetings and the intervening online and telephone discussions. The Panel believes that a centralized delivery system is essential to ensuring the provision of consistent, high quality accessible materials to blind, low vision and print-disabled students.

A centralized approach would:

• Provide an efficient method for validating the integrity and quality of publisher provided baseline NIMAS files.
• Provide for secure archiving of materials protected by intellectual property agreements, thus minimizing the potential for copyright violation.
• Provide a single point of contact for state and local education agencies.
• Increase the potential for nationwide adoption of NIMAS-compliant files.
• Guarantee equal access to instructional materials regardless of a state's technological and materials delivery capacity.

The lack of a centralized distribution system would require individual publishers, conversion houses, and state and local education agencies to each assume responsibility for ensuring the validity and integrity of NIMAS files, the maintenance of copyright compliance, and the implementation of security. An uncoordinated effort of this scope could mean great inefficiency, increased potential for copyright abuse, and wide differences in the quality of accessible materials, their timely delivery, and their availability in the nation's classrooms.

Finally, the development of a centralized distribution system holds considerable promise for eliminating the persistent redundancy created by the existing need to tag or mark up the same set of files twice. As referenced in previous sections, files provided by publishers would contain baseline NIMAS 1.0 markup, and the third party authorized entities would then add DTBook elements from the optional element set. A secure, single point of distribution offers the promise of widespread availability of fully marked up files ready for transformation into student ready versions.

F. Maintenance of the NIMAS Specification

Advances in technology and curriculum publishing will combine with increased awareness in the educational community of the pedagogical benefits of flexible digital materials. These factors alone will suggest modifications and adjustments to the NIMAS specification. In addition, since the NIMAS specification is an application of ANSI/NISO Z39.86 and that latter standard is itself in a continuous maintenance cycle, regular and skilled analyses of the needs of the NIMAS specification with respect to the evolution of ANSI/NISO Z39.86 will be required.

The Technical Panel unanimously recommends that support be provided to a small (10 to 20-member) group of representatives from publishing, conversion entities, education and disability to determine maintenance protocol and versioning policies as the need for them evolves.

V. Additional Recommendations

A. Timeframe for Implementation

Proposed legislation that has incorporated or referenced a NIMAS (The Instructional Materials Accessibility Act; The Individuals with Disabilities Education Act) has consistently projected a two-year phase-in from the date of authorization. In the case of legislated compliance, authorization equals enactment of the legislation itself. In the case of voluntary compliance, authorization represents the formal designation of the NIMAS by the United States Department of Education. Thus, the Technical Panel recommends a two-year phase-in of the NIMAS specification, with voluntary delivery as soon as possible.

B. Improving Awareness

From the beginning of the NIMAS Technical Panel initiative it has been very clear that the national blind and low vision advocacy organizations have accumulated significant experience and expertise in their understanding of the benefits of accessible digital materials for their constituents. Unified within the framework of the American Foundation for the Blind's AFB Textbooks and Instructional Materials Solutions Forum, these groups are responsible for sustaining the advocacy and the technological expertise behind the Instructional Materials Accessibility Act of 2002 (IMAA), proposed legislation that first identified the need for a NIMAS. These organizations have been working for many years to enact legislation guaranteeing the timely availability of high quality accessible materials, and the deliberations of the NIMAS Technical Panel have been enriched and magnified by their previous and their present efforts.

The willingness of representatives of all of the major blind and low vision organizations not only to be actively represented on the Panel but to assist representatives from other disability organizations in understanding the importance of digital media cannot be understated. They repeatedly shared their accumulated wisdom, experience, frustration and empathy in an effort to assist other Technical Panel members in avoiding mistakes that they had already negotiated.

The depth of their understanding of the issues involved in acquiring accessible alternate textbooks underscored, in contrast, the relative inexperience of some of the other disability groups with respect to the benefits that those materials offered to their constituent students. This unevenness of experience and awareness highlighted the need to increase the expertise of representatives of other disability groups as to the considerable benefits of accessible alternate format textbooks and instructional materials.

The Panel unanimously agreed that the awareness level and the expertise pertaining to accessible alternate textbooks needed to be expanded within all disability groups in order for these organizations to promote knowledgeable advocacy in support of the NIMAS standard.

C. All Students with Disabilities

As was referenced previously, the Technical Panel recognizes that the provision of publisher files that conform to the NIMAS, while a significant and necessary first step towards a comprehensive accessibility solution, will not meet the accessibility needs of all students with disabilities.

1. Copyright Constraints

Under the current system, only those students with qualified print disabilities may be provided with accessible versions of print materials that have been converted to braille, audio, or digital text without directly seeking permission from (and giving compensation to) the copyright holder. Ironically, the Section 121 copyright exemption (Chafee Amendment), which was designed to ease the difficulties associated with providing students with print disabilities accessible and appropriate versions in alternate formats, emerges as a barrier when applied as a foundation element in an initiative designed to address the needs of all students with disabilities.

If the most conservative interpretation of the Chafee Amendment guidelines are applied, less than 5 percent of the nearly 6 million students who receive IDEA services and support would qualify to receive accessible instructional materials. In practice, there has been some flexibility in interpreting the Chafee Amendment guidelines. Nevertheless, the fact that the Chafee Amendment provides the only legal means of distributing accessible versions of proprietary materials to students with disabilities creates a significant hurdle when it becomes the basis for a broad-based initiative like the NIMAS.

Within the language including NIMAS as a mandate in the current IDEA reauthorization is a proposed modification to the copyright exemption that would allow publishers to provide NIMAS-compliant digital files directly to a centralized repository, conversion entities, or state or local education agencies, specifically for the subsequent provision of accessible versions of textbooks to students with qualifying disabilities. While this modification does not expand access to these materials to a broader range of students than is presently identified under existing copyright law, allowing curriculum publishers to act as authorized entities would significantly facilitate the flow of textbook files from producers to converters to users.

Beyond expanding the existing copyright exemption to include a broader number of content developers or a wider array of eligible students, it is clear that the original Section 121 copyright exemption was meant to function only as a relief valve for individuals for whom print information (including textbooks) was inappropriate. The exemption was not intended to be used as the foundation of a large scale content distribution enterprise, and attempts to make it so do a disservice to everyone involved: individuals with disabilities, intellectual property holders and publishers. Clearly a more flexible and wider-reaching alternative is necessary.

2. Free Market Distribution of Accessible Versions

The most logical and, ultimately, effective approach for the provision of accessible digital versions of print textbooks is the creation of free market distribution model. One such resolution is for publishers to create alternative and accessible versions of print textbooks for direct distribution to states, districts, schools and students at the same time that the print versions are available. Another possible solution is to improve, simplify, and expedite the process for third parties to obtain permission from, and give appropriate compensation to, the textbook copyright holders to make and distribute accessible versions, known in copyright terms as "derivative works." Under both of these approaches, the accessible digital versions could serve a broad range of students needs, beyond the needs identified in the narrow exemption from copyright laws granted by Chafee. The free market approach could eliminate the delay in the development and distribution of accessible digital versions that is now the norm, and the likely delays in distribution that continue to be associated with narrowly focused Chafee-compliant distribution. In order for a free market system to work, however, a number of conditions must be established.

First, education consumers (states, districts, schools) must demonstrate a willingness to pay for the value represented in the publishers' production and delivery of fully accessible instructional materials. If the growing array of state textbook adoption legislation—those mandating the provision of accessible versions that go beyond requiring digital text files to the inclusion of graphi