【24h】

Remembering Roy Dyckhoff

机译:记住罗伊·迪克霍夫

获取原文

摘要

Roy Dyckhoff left us after a long illness in August 2018. Many of us have known him as a teacher, colleague, mentor, friend, collaborator, and coauthor. He is much missed in the academic world, and especially in the Tableaux community, community, of which he was a founding father and an extremely active member. We shall remember Roy as a scientist with a broad range of interests, care for rigour, passionate approach to new ideas, and enthusiasm for new projects. He showed a human approach to scientific endeavour, had great care for acknowledging priorities, was generous in helping others and did not spare his personal involvement in easing conflicts and striving for justice. In his early years as a researcher, those of doctoral and postdoctoral study, Roy Dyckhoff gave substantial contributions to topology and category theory, the latter also studied in relation to Martin-Loef's type theory. Moving from mathematics to computer science he became interested in computational logic. In the early 1990s he started a systematic study of the use of sequent calculus as a basis for automated deduction, his most influential discovery being a terminating sequent calculus for intuitionistic propositional logic, known as G4 and published in 1992. He was not content with just any solution, but was always looking for the most elegant one. So he returned recently to this issue, improving the proof of the main result. Intuitionistic logic was a main thread of his research; to use his words, he surveyed "the wide range of proof systems proposed for intuitionistic logic, emphasising the differences and their design for different purposes, ranging from ease of philosophical or other semantic justification through programming language semantics to automated reasoning" as well as decision procedures and implementations thereof. In investigating the relationship between natural deduction and sequent calculus he settled an old problem on the relationship between cut elimination, substitution and normalisation. Furthermore, he studied the translations from intermediate logics to their modal companions as well as to the provability logic of Grzegorczyk logic, thus offering a fresh proof theoretic perspective on earlier semantical results. By his contributions relating sequent calculus and natural deduction he shed light on the connections between logic programming and functional programming, for instance regarding the concept of uniform proof. Roy Dyckhoff was appealed by the use of term rewriting techniques in proof theory, and explored innovative extensions of the Curry-Howard-De Bruijn correspondence, which relates formulae to types and proofs to functional programs. He contributed to the development of proof-term grammars and typing systems corresponding to various sequent calculi, with the notion of cut giving a natural typing rule for explicit substitutions, and with cut-elimination being expressed as terminating proof-term normalisation procedures. His contributions to this approach involve for instance the focussed sequent calculi LJT and LJQ, the sequent calculus G4, and Pure Type Systems. Roy Dyckhoff gave important contributions to the method of "axioms as rules"; in particular he proposed a view of rules as rewrite conditions and applied it to obtain a simple decision method, based on terminating proof search in a suitable sequent calculus, for the fragment of positively quantified formulas of the first-order theory of linearly ordered Heyting algebras. Recent work broadens the range of applications of the methodology of "axioms-as-rules." Not only many interesting mathematical theories can be expressed by means of coherent/geometric implications, classes of axioms that can be easily turned into rules, but any first order theory is amenable to such a treatment insofar it has a coherent conservative extension. Often classical conversion steps, such as those based on conjunctive and disjunctive normal form can (and should) be avoided. For this purpose, he devised a new algorithm of "coherentization" that preserves as much as possible of the formula structure. Roy Dyckhoff investigated proof theory also from the more general perspective of proof-theoretic semantics, in particular various notions of harmony, the question of what it is to be a logical constant, favouring the view that leads to strong normalisation results, and the relationship between general and "flattened" elimination rules. He also developed a proof-theoretic semantics for a fragment of natural language as an alternative to the traditional model-theoretic semantics. His scientific interests included systems of multimodal logics for encoding and reasoning about information and misinformation in multi-agent systems. For such logics he employed nested sequent calculi, a formalism beyond traditional Gentzen sequents and gave a Prolog implementation of a decision procedure. Roy Dyckhoff has always been fascinated by the challenge of understanding the classics by modern means, as he did for Frege's Begriffsschrift notation. More recently, he was engaged with Stoic logic. In he showed that the extension of the Hertz-Gentzen Systems of 1933 (without thinning) by a rule and certain Stoic axioms preserves analyticity, which in turn yields decidability of propositional Stoic logic. His latest publication shows how the rule of indirect proof, in the form with no multiple or vacuous discharges used by Aristotle, may be dispensed with, in a system comprising four basic rules of subalternation or conversion and six basic syllogisms. As is clear from this necessarily incomplete summary, Roy Dyckhoff had a proactive attitude that fostered collaboration. He always took genuine interest in the work of others. Many researchers across computer science, mathematics, and philosophy profited from his wide knowledge, deep insights and open-minded approach. He was exemplary in giving credit to others rather than claiming it for himself, and in setting high standards, while at the same time being gracious to those who did not meet them. His humility and his approach to academic life will continue to be an inspiration to all.
机译:罗伊·迪克霍夫(Roy Dyckhoff)在2018年8月患了长期病后离开了我们。我们许多人都将他称为老师,同事,导师,朋友,合作者和合著者。在学术界,尤其是Tableaux社区,他是一位创始之父,也是一位非常活跃的成员,对此他非常怀念。我们将记住罗伊(Roy),他是一位兴趣广泛的科学家,他对严谨,对新思想的热情对待以及对新项目的热情充满热情。他表现出一种人性化的科学努力方法,非常重视承认优先事项,慷慨地帮助他人,并且不遗余力地参与缓解冲突和争取正义的工作。罗伊·迪克霍夫(Roy Dyckhoff)在其博士和博士后研究的初期,为拓扑和范畴论做出了重大贡献,后者也与马丁·洛夫的范畴论进行了研究。从数学到计算机科学,他对计算逻辑产生了兴趣。在1990年代初期,他开始系统地研究使用顺序演算作为自动演绎的基础,他最有影响的发现是直觉命题逻辑的终止顺序演算,被称为G4,于1992年出版。任何解决方案,但一直在寻找最优雅的解决方案。因此,他最近又回到了这个问题,改进了主要结果的证明。直觉逻辑是他研究的主要内容。用他的话来说,他调查了“针对直觉逻辑提出的各种证明系统,强调了差异和针对不同目的的设计,从通过编程语言语义学的哲学或其他语义学证明的易用性到自动推理”以及决策程序及其实现。在研究自然演绎和后续演算之间的关系时,他解决了割消除,替代和规范化之间关系的一个老问题。此外,他研究了从中间逻辑到模态伴随物以及Grzegorczyk逻辑的可证明性逻辑的翻译,从而为早期的语义结果提供了崭新的证明理论观点。通过他对顺序演算和自然演绎的贡献,他阐明了逻辑编程和函数编程之间的联系,例如关于统一证明的概念。罗伊·迪克霍夫(Roy Dyckhoff)在证明理论中使用术语重写技术而颇受人们的青睐,并探索了Curry-Howard-De Bruijn函式的创新扩展,该函式将公式与类型相关,并将证明与功能程序相关联。他为证明项语法和与各种后续演算相对应的分型系统的开发做出了贡献,其中cut的概念为显式替换提供了自然的分型规则,而cut消除则表示为终止证明项规范化过程。他对这种方法的贡献包括,例如,聚焦后继结石LJT和LJQ,后继演算G4和Pure Type系统。罗伊·迪克霍夫(Roy Dyckhoff)为“以公理为规则”的方法做出了重要贡献。特别是,他提出了一种将规则视为重写条件的观点,并将其应用于在适当的顺序演算中基于终止证明搜索的基础上,针对线性有序Heyting代数一阶理论的正量化公式的片段获得一种简单的决策方法。最近的工作拓宽了“轴心即规则”方法学的应用范围。不仅许多有趣的数学理论都可以通过相干/几何含义,可以轻松转变为规则的公理类别来表达,而且任何一阶理论都可以在具有连贯的保守扩展性的情况下接受这种处理。通常可以避免经典转换步骤,例如基于合取和析取范式的转换步骤。为此,他设计了一种新的“相干化”算法,该算法尽可能保留了公式结构。罗伊·迪克霍夫(Roy Dyckhoff)还从更广义的证明理论语义学角度研究了证明理论,特别是各种和声概念,什么是逻辑常数的问题,主张导致强烈归一化结果的观点以及两者之间的关系。一般和“扁平化”的消除规则。他还为自然语言的片段开发了证明理论的语义学,以替代传统的模型理论语义学。他的科学兴趣包括用于在多主体系统中对信息和错误信息进行编码和推理的多模式逻辑系统。对于这种逻辑,他采用了嵌套的后续结石,这是超越传统Gentzen序列的形式主义,并提供了决策程序的Prolog实现。罗伊·迪克霍夫(Roy Dyckhoff)一直对以现代方式理解经典的挑战着迷,就像他为弗雷格(Frege)的Begriffsschrift表示法所做的那样。最近,他从事Stoic逻辑。他在书中指出,通过规则和某些史托克公理来扩展1933年的赫兹-根岑系统(不作稀疏)可以保留分析性,这又可以得出命题性史托克逻辑的可判定性。他的最新出版物表明,在包含四个替代或转换的基本规则和六个基本的三段论的系统中,如何免除间接证明规则(没有亚里士多德使用的多重放电或虚空放电)。从这个必然不完整的摘要中可以明显看出,Roy Dyckhoff的积极态度促进了合作。他总是对他人的工作产生真正的兴趣。计算机科学,数学和哲学领域的许多研究人员都受益于他的广泛知识,深刻的见识和开明的方法。他在给别人以荣誉而不是为自己赢得荣誉方面树立了榜样,并树立了高标准,同时对那些没有达到要求的人表现出仁慈。他的谦逊和对学术生活的态度将继续成为所有人的灵感。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号