In 2025, we started work on what was termed the "TRTM Dictionary of Magnonics" (the Dictionary), with the aim of providing a common ground for further research development in magnonics, useful to and usable by both experienced researchers and those just starting work in the field. However, its first version, "TRTM Dictionary of Magnonics 2025", contained only the three key definitions -- those for "magnonics", "magnonic device", and "magnonic crystal". This reflects the scale and difficulty of the challenge when approached rigorously and ultimately leads to the changes described below.
The drastically increased volume and variety of the definitions present in the Dictionary now have resulted from the use of large-language models (LLMs). This represents a major twist in the direction of the Dictionary, one that was perhaps as inevitable as natural and justified. Indeed, compiling a dictionary of terms is a natural task for an LLM, which effectively synthesises and summarises knowledge accumulated across the community over many years of research. The use of artificial intelligence (AI) is balanced by the expertise of the TRTM community, who curate and refine the LLMs' suggestions and therefore retain the ultimate responsibility for the contents of the Dictionary. Not only does this lend credibility to the outcome of the process, but it also effectively provides feedback for the LLMs' training, especially as the use of the Dictionary proliferates through the community and is reflected in the published literature. In this way, the Dictionary is expected to benefit the research community while also contributing to the refinement of LLMs.
As before, the intention is to keep the Dictionary neutral and inclusive. Yet, we depart from the "contribution without attribution" principle, as we hope to stimulate community engagement with the effort. Instead, we introduce the role of the Editor (or Editors) as the person who takes the primary responsibility for the work for each edition of the Dictionary. All other contributors will be listed, or otherwise appropriately acknowledged, in Appendix B, citing (if requested) their individual contributions. Appendix B will also include any appropriate disclaimers. As a rule, the first draft of each edition will be posted on the TRTM web page and distributed through other means. Then, the Dictionary's main contents (i.e. the definitions) will be revised and augmented with appropriate discussion and conclusion sections, in line with the feedback received from the delegates of the TRTM conference (onsite) as well as that from the rest of the community (by correspondence). We will then seek to publish this and subsequent versions of the Dictionary in a suitable form under an appropriate open access licence, currently at Zenodo (https://zenodo.org/). Everyone will be able to use the contents of the Dictionary in their own publications and presentations (and is encouraged to do so), perhaps citing the relevant version and adding a web link to the TRTM 2026 website.
We intend to update and revise the Dictionary annually, in line with the TRTM organisation cycle. The core contents of the Dictionary (although not necessarily the accompanying text) will remain free from references and claims of priority. Whenever there are multiple competing and sufficiently distinct definitions, we will endeavour to list all the definitions. Where applicable (for example, when consensus cannot be reached in time, or when multiple versions exist but are not sufficiently distinct), a summary of the discussion points will be included, albeit kept minimal. The terms and definitions included in one version of the Dictionary will remain open for discussion and further improvement in subsequent cycles of the work.
Briefly, the following methodology was adopted when preparing this version of the Dictionary.
-
First, Microsoft 365 Copilot was used to generate, one by one, definitions for the terms that were listed without definitions in the previous version of the Dictionary while requiring that they be consistent with the three key definitions therein.
-
Second, each definition was individually refined by the Editor with assistance from the Copilot, typically using the prompt "What about [revised definition]?" and then revising the definition in light of the feedback received. Only a small fraction (5-10%) did not require refinements, and about 5 iterations were required on average.
-
Third, Copilot was asked to evaluate the resulting version:" Evaluate the Dictionary from the attached file in terms of completeness, self-consistency, and consistency with published literature on magnonics." The feedback received was used to manually revise and expand the master copy of the Dictionary (to avoid introducing artefacts) to include new terms and harmonise the draft. The process was repeated multiple (20+) times, accounting for most of the suggestions, while refining each revised definition either individually or as a small sub-group of related definitions. The version below corresponds to the stage when the Editor felt that the Dictionary had reached a form suitable for further scrutiny by the wider community.
-
In the final stage of the process, the whole document, including this Foreword and Appendices, was evaluated by Copilot, and final refinements were introduced.
-
Changes introduced to the text other than the definitions and section headings were purely grammatical and, in some instances, stylistic, i.e. Copilot was not asked or used to generate any of the text.
-
Throughout, Copilot was used in the "Think deeper" mode with a custom instruction: "Share clear and truthful feedback in a respectful tone, and focus on offering constructive suggestions when something needs improvement."
-
The procedure was repeated once preliminary feedback had been collected.
-
At the time of starting the process, the Editor had only used Copilot a few times over a couple of weeks, and then for tasks largely unrelated to this work.
The following editorial choices and conventions were made throughout the Dictionary and, although open to discussion, are likely to persist in future versions.
-
"Spin wave(s)" and the adjectival form "spin-wave" are used when the wave character of the excitation, including its phase, propagation, interference, or mode structure, is important.
-
"Magnon(s)" and the adjectival form "magnon" are used when the quasiparticle, quantum, statistical, or transport character of the excitation is important.
-
In situations where both the spin-wave and magnon descriptions are appropriate, either both terms are used, or the broader adjective "magnonic" is used instead. In particular, the phrase "magnonic signal" is used where the signal may be described either as a coherent spin wave or in magnonic terms, or where the distinction is not essential for the definition.
-
The adjective "all-magnonic" is used to emphasise that the effect, phenomenon, or function is produced, mediated, or performed solely by spin waves and magnons, without conversion to other physical carriers, implicitly assuming that external input/output transduction may be present.
-
The adjectives "reconfigurable", "switchable", "tuneable", and "(re)programmable" are used with related but distinct meanings. "Reconfigurable" is used as the broadest term for systems whose magnonic properties or functions can be reversibly modified without structural changes. "Switchable" is used when such modification occurs between two or more distinct states. "Tuneable" is used when the relevant property or function can be adjusted continuously or quasi-continuously by external control parameters. "(Re)programmable" is used when the relevant property or function can be set, erased, and reset, typically according to an externally imposed instruction, protocol, or procedure.
-
The dimensional labels (such as "1D" or "3D") are used to highlight the effective dimensionality of spin-wave propagation or magnon transport. In the context of magnonic crystals, they refer primarily to the number of spatial dimensions in which the relevant magnonic properties vary periodically.
-
The half-dimensional labels (e.g. "1.5D") are used when, in addition to the primary effective dimensionality or periodicity, the relevant magnonic properties or characteristics are also significantly influenced by structure, confinement, nonuniformity, or variation in an additional spatial dimension.
-
The label "4D" is used for systems in which the relevant magnonic properties vary in both space and time.
-
The words "medium", "structure", "system", and "device" are used pragmatically. "Medium" and "structure" emphasise the physical platform supporting spin waves or magnons; "system" is used for a broader physical arrangement or dynamical entity; and "device" is reserved for cases where a specific function or application is central to the definition.
-
The terms "allowed" and "forbidden" are used in relation to bands, band gaps, and propagation in the idealised sense customary for wave systems. In lossy, disordered, finite, or otherwise imperfect systems, "forbidden" may correspond to propagation being strongly suppressed rather than strictly absent.
-
The definitions are intended to state the meaning of terms rather than to provide exhaustive physical descriptions. Qualifying words such as "typically", "often", "usually", and "may" are used where a term has multiple realisations, conventions, or limiting cases.
-
Abbreviations, acronyms, and alternative names are included where they are common in the literature or useful for recognising equivalent terminology. Their inclusion does not necessarily imply a preferred term unless this is clear from the entry.
The following editorial choices were made here but might be revised in future.
-
Copilot suggested assigning each term a label ("established", "emerging", and "proposed") reflecting its maturity, as they "differ not only in subject matter, but also in their degree of consolidation within the published literature. Some belong to the long-established and widely shared vocabulary of spin-wave physics and magnonics; some are clearly present in current specialist literature but are still in the process of broader standardisation; and some are useful and internally consistent extensions whose continued discussion may, in time, lead to their wider adoption." Upon inspection of the allocation and further evaluation, we decided not to do this. Although potentially useful and done following a coherent methodology (suggested by Copilot), the allocation would necessarily be subjective and volatile. Indeed, every user of this Dictionary is likely to have their own opinion on the topic, while still being able to seek expert advice from their colleagues or to use their preferred AI tools to provide the required classification, tailored to their needs. In other words, this does not appear to be a task that could benefit from the centralised curation and editing in the same way as the Dictionary itself.
-
Copilot recommended separating the Dictionary (which tells the reader what a term means) from the taxonomy (which tells the reader how concepts are grouped and related to one another within the field). Specifically, this would mean making a distinction "between dictionary entries, which define terms of established or emerging usage, and taxonomic labels, which organise the field into subfields, subclasses, and related concept families. Taxonomic labels are included to aid navigation and discussion, but they should not necessarily be read as equally established lexical headwords." We decided only to draw the reader’s attention to the fact that such a distinction naturally exists, albeit we do not make it explicit in this version of the Dictionary. As an example, we point out that most of the entries in Section 1 are "true" dictionary terms, while Section 2 de facto contains the taxonomy of magnonics; also, many terms are formed by adding an "organising label" (e.g. "magnonic" or "all-magnonic") to a dictionary term.
-
New terms, of various relevance to magnonics, kept emerging or proposed for adding to the Dictionary. Eventually, it was decided to draw a line under the process for now, adding Appendix A for "Terms considered for inclusion in future editions of the Dictionary", most likely in its 2027 edition. You are welcome to suggest terms for the list, perhaps even with proposed definitions, although such suggestions will be considered selectively.
As a final note in this Foreword, we stress that our approach to the Dictionary remains flexible and will be adjusted as we learn how it works in practice and receive feedback from the community. For now, enjoy the Dictionary and let us know your opinions and suggestions!