In search of an optimal method for analyzing deep structures: frame semantics and classification of argumentative structures

Chilingaryan Kamo Pavelovich ORCID: 0000-0002-3863-8603 Associate professor, Department of Foreign Languages, Higher School of Management at Peoples' Friendship University of Russia named after P. Lumumba 10 Miklukho-Maklaya str., office 707, Moscow, 117198, Russia chilingaryan_kp@pfur.ru Other publications by this author

Sorokina Lyudmila Stanislavovna ORCID: 0009-0006-4656-3214 Senior lecturer, Higher School of Management, RUDN University named after Patrice Lumumba (Foreign languages department) 117198, Russia, Moscow region, Moscow, Miklukho-Maklaya str., 10, office 707 camilla49@mail.ru

Introduction

The relevance of the research is due to different approaches and interpretations to "deep structures". Long enough the point of view of Ch. Fillmore regarding the "deep cases" did not cause controversy, but at the beginning of the new century, some of his postulates were critically interpreted. In the presented work, the authors try to find common ground between the main critics of the "deep cases". The findings can be applied in various fields, and the study of different approaches allows you to compare their effectiveness and determine the most appropriate method for specific tasks. A comparative analysis of approaches to "deep structures" allows researchers to understand the advantages and limitations of each approach, which helps to optimize the use of methods and develop new approaches taking into account the strengths and weaknesses of existing solutions. The semantics of Fillmore frames allow for a more complete understanding of semantic interpretation, covering rich contextual and situational aspects of meaning, and VerbNet Levin and Rappaport provide a more detailed analysis of verb-specific models. Disagreements and alternative points of view contribute to the constant development of linguistic theories and open up new prospects for further research in this field.

In the field of artificial intelligence (AI) and computational linguistics, the approaches of case grammar, deep cases, VerbNet and argument implementation play a crucial role in improving our understanding of language processing and understanding of natural language. These linguistic structures provide a systematic way to analyze syntactic and semantic structures of sentences, helping computers to understand and generate human language more effectively.

The relevance of these approaches in the context of modern technologies cannot be overestimated. In the context of the rapid development of artificial intelligence and the growing dependence on computers for various tasks, the ability to accurately analyze and interpret natural language is essential. By applying the principles of case grammar, deep cases, VerbNet and argument implementation, researchers and developers can improve the accuracy and efficiency of natural language processing systems, making them more capable of understanding the nuances of human communication.

The relevance of this issue is evident in the growing demand for artificial intelligence technologies capable of understanding and generating human language. From virtual assistants such as Siri and Alexa to language translation services and chatbots, there is a clear need for artificial intelligence systems capable of efficiently processing and generating natural language. By incorporating knowledge from case grammar, deep cases, VerbNet ^[12] and argument implementation into these systems, developers can improve their performance, make them more convenient and effective.

For the effective use of these linguistic structures in the field of artificial intelligence and computational linguistics, certain prerequisites must be met. Researchers and developers should have a clear understanding of linguistic theory and computational methods, as well as knowledge of programming languages and machine learning algorithms. In addition, a collaborative approach involving experts in linguistics, computer science and artificial intelligence is needed to ensure the successful integration of these approaches into AI systems.

By providing students with the knowledge and skills to apply these linguistic frameworks to real-world applications, teachers can prepare the next generation of AI professionals to meet the needs of the science and technology industry. Recognizing the relevance of this framework and investing in the necessary research and education, it is possible to meet the needs of the scientific and technical industry and pave the way for the further development of language processing and artificial intelligence technologies.

In this study, the authors set themselves the following tasks: – to study different approaches to deep cases and describe the existing differences in the approaches of two representatives of the cognitive approach (C.Fillmore and B.Levin–M.H.Rappaport), identify significant differences and/or similarities, taking into account, respectively, the points of view on the semantics of frames and Verbnet, analyze and answer the question: are the different points of view of these scientists in contact.

One of the research methods for the analysis of deep cases, frame structure and semantic roles is the semantic analysis of sentences based on corpus data. This method includes the study of various linguistic constructions in the context of their use in real texts, which allows us to identify common patterns and rules for the use of these linguistic units in different situations. Such an analysis can help uncover patterns in the structure of sentences and interpret their semantics. Comparative analysis of the approaches of Ch. Fillmore and Levin-Hovav make it possible to compare similarities and differences in understanding the structure of language. Using these methods allows you to scan language data and identify hidden semantic relationships.

The subject of the research is the search for the optimal method of analyzing deep structures using frame semantics. The study of semantic roles, similarities and differences in the approaches of both C. Fillmore and B. Levin - M. R. Hovav allow for a more detailed and accurate analysis of the sentence structure, identify deep cases and determine semantic relationships between words. The study of these aspects is key to understanding language constructs and their interpretation. The study of various approaches makes it possible to identify both common features and unique features, which is key for a complete understanding of language structures. An interest in text analysis in the field of artificial intelligence, machine learning and computational linguistics, and an understanding of the semantic relationships between words will help create more accurate and efficient text processing algorithms.

The research material was the work of I. V. Bondarenko (2011), Carter, R.J. (1978), Palmer M., Kingsbury P, Gildea D. ^[23], as well as FrameNet electronic resources ^[11] https://framenet.icsi .berkeley.edu, VerbNet. http://verbs.colorado.edu/~mpalmer/projects/verbnet.html, WordNet ^[13]  http://wordnet.princeton.edu

The theoretical basis of the research is the work of such scientists as C. Fillmore (1981, 1986), N. Chomsky (1965), Levin, B. ^[19],[21], Levin, B. and Hovav M.R. (2005), G.F. Lutfullina (2022), V. Vs. Ivanov, (2004), E.A. Krasina (2018), S.V. Ryadinskaya (2006), Labov W. (1973) S.Y. Kravchenko (2015), M.V. Markova (2013). The works of A. Goldberg (2003), A. McIntyre (2005), R. Jackendoff (2002) were used as alternative points of view of scientists. In the work of Lutfullina G.F. ^[5], the question of five typological criteria of case grammar is raised, while V. V. Ivanov ^[2] examines the languages of the third millennium from the point of view of semantics and meanings affecting the structure of cases. S.V. Ryadinskaya ^[7] concludes about a timeless set of semantic and syntactic functions ("roles") for named components. An important aspect of the research was the work of McIntyre and Jackendoff, which raises issues related to conceptual and empirical problems, and Goldberg, who pays attention to the role of constructive patterns. In the work, the authors also refer to the dissertation of one of the authors of this study, K.P. Chilingaryan ^[10], which states that "case is a relational and functional category, since nominal forms of a simple sentence and their main functions are to be a subject or object, being involved in the sphere of the verbal predicate, by their nature, turned out to be functional-semantic, i.e. deep meanings, and superficial formal-grammatical structures, as a rule, accompany deep meanings and structures."

The practical significance of the research lies in its application in various fields, including computer and applied linguistics and psychology. The ideas of grammatical constructions reflect the semantic relationships and interactions between words in a sentence. This can be useful for developing more accurate descriptions of language structures and systematic analysis of word meanings, for automatic translation, speech synthesis, text analysis and other language processing tasks. The work is intended to become a source of information for linguists, teachers and students interested in the role of deep cases in the syntactic structure of the language. Students need a deep understanding of grammar and language structures for effective communication and academic success. Teachers rely on this knowledge to develop meaningful lessons and assessments that meet the needs of their students. Researchers use in-depth case analysis to explore complex phenomena and discover new ideas in their fields of study. In general, the use of these tools and methods plays an important role in meeting the needs of the scientific community and promoting progress in academic research. The analysis of deep cases can be important in various subject areas such as psychology, sociology, education and business.

Discussion and results

Speaking about Fillmore's case grammar ^{[8, pp. 369-495]}, it should be emphasized that it implies a hierarchical organization of deep structures. Fillmore's grammar, known as "case grammar", is one of the key theoretical foundations used in the analysis of deep structures. It is based on the concept of thematic roles, which are determined based on the semantics of the sentence and reflect the interaction between the verb and its actant. Fillmore believes that deep cases are a tool for expressing these thematic roles and reflect the structure of events in a sentence.

As G.F. Lutfullina notes (Lutfullina, 2022, p. 2376), in the work of Ch. Fillmore considers "the application of five typological criteria of case grammar, namely: the presence of special forms and coordination, taking into account anaphoric processes and processes of communicative selection, including the possibility of choosing the order of words." According to M. A. Rybakov (2012, p. 39), in modern English, semantic roles and/or functions within a sentence manifest themselves as deep cases. The Russian language is dominated by a generally accepted point of view with the allocation of six cases.

In turn, Beth Levin and Malka Rappaport Hovav ^[20] propose a more detailed and comprehensive approach to the analysis of deep structures. They take into account not only the semantics of the sentence, but also contextual and lexico-semantic features. Their concept includes the concept of a "prototypical argument", which is the most significant element in a sentence and defines its deep structure. The prototypical approach is described in more detail in the article by S. Y. Kravchenko (2015). As S. Y. Kravchenko notes ^{[3, pp.117-119]}, referring to the work of V. Labov ^{[18, p. 342]}, "the process of categorization"is such a fundamental and obvious fact of linguistic activity" that the properties of the category themselves are usually taken for granted and therefore not investigated.

The subject-object organization of the sentence of nominative languages, as defined by Fillmore (1981, p. 430), is based on case values and functions of the subject and object. Vyacheslav Ivanov notes that "In most languages where there are cases, they denote the main actants (the relationship between the predicate verb and its arguments - the subject, different objects, the addressee, etc.)" (Ivanov, 2004, p. 79). A similar point of view is shared by Ch. Fillmore ^{[15, p. 163-82]}.

The theory of deep structures still requires proof, since it is connected with other "matters". N. Chomsky ^{[9, pp. 465-576]}, the founder of generative linguistics, as opposed to descriptive, is based on the description of language in the form of formal models of a certain type. The initial type of the latter are generative grammars. Every native speaker uses generative grammar to use it to construct a statement or understand it. According to I.V. Bondarenko ^{[1, pp. 141-149]}, "language competence and language use are essential for generative linguistics."

The deep case in linguistics, developed by C. Fillmore, is a category of semantic kind used to describe the relationship of depth or nesting between subject and object in a sentence. The deep case indicates that the subject of the sentence is more deeply or "intimately" connected with the action or state expressed by the verb than the object. This relationship is expressed using specific syntactic and semantic features.

Examples of the use of the deep case:

1. "Mom baked a cake for the kids." - Here, the subject "mom" is more deeply connected with the action of "baking a cake" than the object "children", since "mom" performs the action directly, and "children" are the recipients of the results of this action.

2. "The doctor performed surgery on the patient." - In this sentence, the subject "doctor" is associated with the action "performed the operation" much more deeply than the object "patient", since the "doctor" performs the action, and the "patient" is the recipient of this action. (The authors compiled these examples using the example of the frame grammar of Ch . Fillmore).

The deep case helps to describe the relationship between parts of a sentence and to understand how various aspects of semantics and syntax interact with each other. It is an important tool in the framework of cognitive language research.

As noted by Ryadinskaya S.V. and Migachev V.A. (2006, pp. 181-183), the works of C. Fillmore originated "case grammar" or role grammar, which is a method of "describing the semantics of a sentence as a system of semantic valences through the connections of the "main verb" with the roles dictated by the meaning of this verb and the performed nominal components." Each case has its own verbs, or each verb corresponds to a certain set of cases, and each of the cases corresponds to a certain participant in the situation.

The Fillmore generative FrameNet model is a linguistic semantic model based on the concept of a "frame", that is, a structure representing organized knowledge about a concept or situation and the associated roles of participants. It consists of slots that are filled with semantic roles and arguments. This model is based on the idea that the semantics of a word or expression is determined by the context in which it is used. The model builds grammatical and semantic links between frames and phrases, allowing you to automatically extract semantic information from texts.

The case grammar of Ch . Fillmore is a model of grammatical analysis that is considered an extension of traditional nominative grammar and offers a more detailed description of the syntactic structure of a sentence.

The Fillmore model generation process includes the following steps:

Data collection: Large volumes of texts are collected, in which frames and their attributes are automatically highlighted.

Frame annotation: Experts manually annotate semantic frames and their attributes in texts.

Building a dictionary: A dictionary is created that links words and expressions to the corresponding frames and meanings.

Creating grammatical patterns: Defines grammatical patterns that associate frames with the corresponding words and expressions.

Testing and tuning the model: The model is tested on new data and adjusted based on the results obtained.

Fillmore developed a system of seven cases that reflect the semantic roles of various arguments in the frame. These cases include:

A0 (agentive) - expresses the agent or initiator of the action.

A1 (patient) - expresses the patient or the object on which the action is performed.

A2 (recipient/profit) - expresses the recipient or recipient of the action.

A3 (effective/instrumental) - expresses the instrument or means by which an action is performed.

A4 (stylistic) - expresses the place or context in which the action takes place.

A5 (emergency) - expresses an aspect of an action related to time or circumstances.

AM (modifier) - expresses additional information about the action or its participants.

Fillmore's method of case grammar helps to analyze the structure of sentences at the semantic level and allows you to more accurately determine the roles and relationships between its components. This model is actively used in semantic analysis and computational linguistics. Unlike the traditional case system of the Russian language, the Fillmore case system is based on a semantic approach, that is, cases are determined by meaning, and not by grammatical rules.

Comparing the linguistic approaches of Fillmore and his supporters, on the one hand, and Levin and Rappaport, on the other hand, we come across two different but complementary views on language analysis. Fillmore's frame semantics states that words and phrases evoke mental structures or "frames" that help interpret meaning. This approach strongly emphasizes the context and the role of frame activation in understanding language.

Conversely, Levin's VerbNet approach ^[19] focuses specifically on verbs and their syntactic models. By classifying verbs into semantic groups based on their behavior, VerbNet provides a complete understanding of how verbs are used in various contexts. While Fillmore's "Semantics of Frames" offers a broader view of semantic interpretation, Levin's VerbNet offers a detailed analysis of the behavior of verbs. Together, these approaches contribute to a holistic and detailed understanding of language by answering various research questions.

To. Schuler ^[24] emphasizes that VerbNet is a lexical resource that provides information about the syntactic and semantic behavior of English verbs. Its purpose is to cover the different ways verbs are used in sentences, including their argument structure and thematic role. This resource is linked to other linguistic resources such as PropBank. PropBank is a corpus—based resource that links individual instances of verbs in a text with corresponding semantic "frames" (frames). It provides specific information about the roles that verb arguments can play in a sentence. This term refers to the specific corpus of M. Palmer (Palmer et al., 2005). PropBank was formed by combining two words - annotated with verbal sentences and their arguments - "proposal bank". PropBank is focused on verbs, whereas FrameNet is based on the more abstract concept of frames, which generalizes descriptions for similar verbs (for example, "describe" and "characterize"), as well as nouns and other words (for example, "description"). According to SCISPACE (https://typeset.io/questions/what-is-prop-bank-4j89vfzl49 ), PropBank is a manually annotated corpus used for semantic role labeling (SRL), which provides information about predicates and arguments for the language. It includes non-verbal predicates such as adjectives, prepositions, and verbose expressions, and covers a wide range of fields, genres, and languages. The purpose of PropBank is to capture the semantic role of arguments in a sentence, which helps to understand the meaning and relationships between words. To create and update PropBank's frameset files, a special Cornerstone editor is used, which simplifies the process and supports multiple languages. Claire Bonial et al., in the work of PropBank Annotation Guidelines (https://www.researchgate.net/publication/266177244_PropBank_Annotation_Guidelines, 2010) clarify that PropBank is a corpus in which the arguments of each predicate are annotated by their semantic roles in relation to the predicate. In addition to semantic role annotations, PropBank annotations require the selection of a senseid (also known as a "frame set" or role set identifier -rolesetid) for each predicate.

By combining information from the FrameNet, VerbNet, and PropBank classes, researchers can better understand how verbs are used in natural language and how their semantic and syntactic properties relate to each other.

Beth Levine and Malka Rappaport Hovav have made significant contributions to the field of argument implementation. They have developed a theoretical framework that explores how verbs and related arguments are structured in sentences.

In their work, Levin and Rappaport Hovav (2005) suggest that the meaning of a verb determines the types of arguments it requires and how they are implemented syntactically. They claim that verbs have a set of thematic roles or "thematic grids" (theta grids) that represent potential arguments related to the verb. These roles include, among others, agents, patients, tools and place. M.V. Markova in her research ^[6] argues that "there is a correlation between the change in the meaning of the verb and its syntactic behavior. <...> The control model sets the word order in a sentence, which is not taken into account in the theory of frame semantics. The theory of the implementation of arguments by B. Levin and M.R. Hovav emphasizes the interaction between the meaning of the verb and syntactic structures that allow expressing these meanings. They explore how different verbs have different requirements for the number and type of arguments they can take, as well as the syntactic constructions in which they can appear.

Their research provides valuable information about the nature of argument structure and has influenced various fields of linguistics, including syntax, semantics, and computational linguistics. By exploring how verbs and their arguments combine, this work sheds light on the fundamental processes underlying speech formation and language understanding.

Despite the fact that Ch. Fillmore, and B. Levin and M. Rappaport Howav made significant contributions to the study of the implementation of arguments and the role of verbs in the structure of sentences, there are some differences in approaches between them.

Fillmore's Case Grammar (Fillmore, 1981) focuses on the grammatical cases associated with noun phrases and how they relate to the verb. Fillmore's theory emphasizes the importance of syntactic and semantic roles, and he proposed the idea of "deep cases", which represent the basic semantic relationship between a verb and its arguments.

The approach of B. Levin and M. R. Hovav is based on the grammar of Fillmore cases, but also includes additional ideas. Next, they analyze the correspondence between the thematic roles associated with verbs and the grammatical realizations of these roles. Their theory takes into account the various syntactic constructions in which verbs can appear, and how these constructions affect the implementation of arguments.

The work of B. Levin and R. Hovav also focuses on the meaning of the verb and the conceptual structure underlying the implementation of arguments. They argue that the meaning of a verb plays a crucial role in determining the syntactic and semantic properties of the arguments associated with it.

In the classification of verbs according to Levin (Levin, 1993), the following main classes are distinguished:

1) Verbs of Motion: describe moving or changing location (for example, walking, running, flying).

2) Cognitive Verbs: refer to the processes of cognition, thinking and perception (for example, to think, to know, to imagine).

3) Verbs of transition (Transitional Verbs): denote the physical transformation of one object into another (for example, to transform, transform).

4) Verbs of benevolence (Verbs of Welfare): describe actions aimed at improving the well-being of other people or animals (for example, to help, to help, to provide).

5) Verbs of Communication: they are associated with the transmission of information and the expression of thoughts (for example, to speak, to ask, to tell).

6) Verbs of Perception: describe the processes of perceiving the world through the senses (for example, to see, hear, feel).

7) Verbs of Change of State: denote changes in the state of an object, such as appearance, disappearance or transformation (for example, to become, to appear, to disappear).

Levin's classification of verb classes is an attempt to describe various semantic types of verbs, their basic meanings and related features. It helps to structure and systematize lexical material within the framework of lexicographic and linguistic research.

 We conclude that although both Fillmore and Levin and Rappaport Howav contributed to our understanding of the implementation of arguments, the latter's approach is based on Fillmore's work and includes additional ideas, especially regarding the meaning of the verb and the interaction between meaning and syntax.

Within the framework of case grammar, Ch. Fillmore's "case" refers to the grammatical labeling or categorization of noun phrases (NP) based on their syntactic and semantic role in a sentence. Fillmore's understanding of case differs from the traditional concepts of case found in other linguistic systems, such as nominative, accusative, or dative cases in traditional Latin or German grammar.

According to Fillmore, the case is determined not only by the morphological form of the nominal group, but also by its function in a particular sentence. He proposed the concept of "deep cases", which represents the basic semantic relationship between a verb and its arguments. These deep cases can then be implemented in various grammatical forms, such as prepositions, word order, or other syntactic constructions. The author of the article believes that the terms "superficial" and "deep" cases distinguish the level of realization of the form as a phenomenon and the level of competence (interpretation) of the content as an entity. "<...> case is a relational and functional category, since the nominal forms of a simple sentence and their main functions are to be a subject or object, and being involved in the sphere of the verbal predicate, by their nature, turned out to be functionally semantic, i.e., deep meanings, and superficial formal grammatical structures, as a rule, they are accompanied by deep meanings and structures." (Chilingaryan, 2022, p. 159). E.A. Krasina ^{[4, p. 126]} mentions that "modern methods and theories developed on their basis are based on the principle of integral description of language, taking into account the interaction of the expression plan and the content plan <...>; deep (semantic) and surface (syntactic) structures of N. Chomsky.

Fillmore identified a set of basic deep cases that correspond to the various thematic roles associated with verbs. These deep cases include, among others, agent, patient, expert, subject, and instrument. The choice of the deep case for a particular argument depends on the meaning of the verb and the semantic role that the argument performs.

For example, in the sentence "John kicked the ball", "John" will be associated with the deep case of the agent, and "ball" with the deep case of the patient. The specific implementation of these deep cases may vary depending on the language and its grammatical structure.

The structure of Fillmore's case grammar makes it possible to analyze the relationship between verbs and their arguments based on their semantic role. She emphasizes the idea that the choice of case is not determined solely by morphological inflection (change), but is motivated by the basic meaning and structure of the sentence.

If the structure of the case grammar is Ch . Fillmore focuses primarily on the concept of deep cases and their implementation in various grammatical forms, B. Levin and M. R. Hovav expanded this idea and further developed the understanding of case.

Levin and Rappaport Hovav expanded the case analysis to include a more detailed study of the meaning of the verb and its impact on the implementation of arguments. They proposed the concept of "thematic grids" or "thematic roles", which represent potential arguments related to the verb and their specific syntactic and semantic properties.

In their work, Levin and Rappaport Hovav investigated why different verbs have different requirements for the number and type of arguments adjacent to them. They examined the specific syntactic constructions in which verbs appear, and how these constructions affect the case implementation of their arguments.

In addition, they explored the interaction between verb meaning, thematic grids, and syntactic and semantic properties of arguments. Their analysis took into account not only the traditional concepts of case marking, but also other syntactic features and constructions that contribute to the implementation of the argument, such as prepositions, word order and syntactic alternation.

By incorporating a more comprehensive approach to case, Levin and Hovav provided a deeper understanding of how case is determined both by the meaning of the verb and by the syntactic structures that allow these meanings to be expressed. Their work expanded the scope of case analysis and contributed to our understanding of the implementation of arguments in a broader sense.

In their work, B. Levin and M. R. Hovav relied on the ideas of Ch. Fillmore talked about verb classes and expanded the analysis by presenting a cross-classification structure that takes into account both the meaning of the verb and the structure of the arguments.

Like Fillmore, Levine and Rappaport Howav recognized that verbs can be divided into classes based on their common properties. However, Levin and Hovav went further by proposing a more detailed classification system that takes into account not only the thematic roles associated with verbs, but also the syntactic structures in which these roles are implemented.

Levin and R. Hovav's cross-classification system includes the classification of verbs into more specific classes based on the number and type of arguments they take, as well as the syntactic constructions and alternations in which they participate. They investigate the diverse behavior of verbs in various argument structures and identify patterns of implementation of arguments specific to specific classes of verbs.

This approach differs from Fillmore's verb classes, which were primarily based on semantic roles and deep cases. The Levin and Hovav cross-classification system combines both semantic and syntactic features, providing a more complete analysis of the behavior of verbs and the structure of arguments.

By including a cross-classification of verbs, the work of Levin and Hovav allows us to understand in more detail the relationship between the meaning of the verb, the structure of arguments and syntactic alternations. It emphasizes the variability and flexibility of the implementation of verbs in various syntactic constructions, enriching our understanding of the complex interaction between verbs, arguments and sentence structure.

Referring to the work of Carter ^{[14, pp. 61-92]} and Wilkes ^{[25, pp. 759-761]} B. Levin (2009) says that "the meanings of verbs represent interpretations of events," which is really different from Fillmore's approach, although there are some overlapping ideas.

Within the framework of case grammar, Ch. Fillmore focuses on the relationship between the meanings of verbs and the structure of sentence arguments. Fillmore emphasizes the role of deep cases and their implementation in various syntactic constructions. Although Fillmore was aware of the importance of the verb's meaning in determining the structure of an argument, his main focus was on the grammatical and syntactic aspects of case realization.

On the other hand, the idea that "verb meanings represent interpretations of events" highlights a broader view of how verb meanings relate to the conceptual representation of events. This concept, often associated with the works of Levin and Hovav, suggests that the meanings of verbs not only determine the syntactic structure, but also form the conceptualization or interpretation of events.

Levin and Hovav argue that the meanings of verbs involve a process of interpretation that reflects how events are mentally presented and perceived. They explore how different verbs can highlight or emphasize certain aspects of an event, such as the method, outcome, or participants. This interpretive perspective examines the cognitive and interpretative aspects of verb meanings beyond their syntactic realization.

Thus, although Fillmore's work focuses more on the grammatical and syntactic aspects of case realization, the notion that "verb meanings represent interpretations of events" expands the perspective to include cognitive and interpretative aspects of verb meanings and their impact on the conceptualization of events.

There are often healthy debates and discussions in linguistics around various theories and concepts, including the work of Beth Levine and Malky Rappaport Howav. Along with the very important work of Levin and Hovav, there are other researchers with other points of view and alternative proposals. Here are some examples of scientists who have expressed alternative points of view or criticism:

Ray Jackendoff ^[16] expressed some disagreement with Levin and Hovav's approach to the implementation of arguments. He advocated a different theoretical framework that emphasizes the role of conceptual structure and conceptual semantics, which differs from the point of view of Levin and Hovav about thematic grids specific to verbs.

Ray Jackendoff's point of view can be seen as being somewhere between the ideas of Levin and Rappaport Howav and Ch. Fillmore, as she incorporates elements of both approaches while presenting her own ideas.

Jackendoff, like Levin and Rappaport Hovav, emphasizes the role of conceptual structure in the implementation of arguments. He emphasizes the importance of considering the conceptual representations and semantic roles associated with verbs and their arguments. This is consistent with Levin and Hovav's idea of the meanings of verbs and the interpretation of events.

However, Jackendoff also notes elements of Fillmore's case grammar, in particular, deep cases and their implementation in various syntactic constructions. He examines the relationship between argument structure and syntactic structure, recognizing the influence of both conceptual and grammatical factors.

To sum up, Ray Jackendoff's point of view can be seen as a combination of Levine's ideas/ Rappaport Howav and Fillmore. It includes elements of the conceptual structure and the meaning of the argument, as well as considers the syntactic and grammatical aspects of case realization.

Andrew McIntyre ^[22] challenges the conceptual and empirical approaches of Levin and Rappaport Howav to the structure of argumentation. He advocated a more flexible model that takes into account a wider range of factors, such as context and discourse, when determining the structure of arguments, rather than relying solely on thematic grids specific to verbs.

Andrew McIntyre's point of view can be seen as closer to Levin and Rappaport Howav's approach than to Fillmore's, although he criticizes some aspects of Levin and Howav's work while agreeing with their overall structure.

McIntyre raises questions related to the conceptual and empirical problems of Levin and Hovav's approach to the structure of argumentation. He advocates a more flexible model that takes into account a wider range of factors, such as context and discourse, when determining the structure of an argument.

Although McIntyre criticizes some aspects of Levin and Hovav's approach, such as reliance on specific thematic grids specific to verbs, his own proposals still follow the outline of a broader framework for considering the meaning of the verb and contextual factors that form the structure of the argument. This coincides with Levin and Rappaport Howav's view of the interaction between verb meaning, argument structure, and event conceptualization.

On the contrary, Fillmore's case grammar structure pays more attention to the grammatical and syntactic aspects of case realization, which differs from McIntyre's call for a more flexible and context-sensitive approach.

Thus, McIntyre's point of view is the closest to the ideas of Levin and R. Hovav, since he, using their structure, offers alternative ideas.

Adele Goldberg ^[17] offered different views on the structure of arguments and the meaning of verbs, focusing on the role of constructive patterns and how the structure of arguments arises as a result of the interaction between verbs and constructions. Her work presents a point of view different from Levin and Hovav on thematic grids specific to verbs.

Adele Goldberg's point of view can be seen as closer to Fillmore's approach, as it shares some common features with his ideas, but at the same time presents its own point of view.

Goldberg's work highlights the role of constructive patterns in shaping the structure of argumentation. She argues that the structure of arguments arises from the interaction between verbs and the linguistic constructions in which they appear. This is consistent with Fillmore's special attention to the syntactic and grammatical aspects of case realization.

Like Fillmore, Goldberg recognizes the importance of both verb meaning and syntactic constructions in determining the structure of arguments. She explores how verb-specific constructions and constructive schemes affect the implementation of arguments.

However, Goldberg's approach also introduces some new concepts and ideas. She emphasizes the importance of considering productivity and generalization of constructions, as well as the role of constructive meaning in the implementation of an argument. Her work extends the understanding of the structure of arguments beyond the individual properties specific to verbs and includes broader generalizations about the interaction between verbs and constructions.

Thus, although Adele Goldberg's point of view has some similarities with Fillmore's, especially in considering the role of syntactic constructions in the structure of arguments, she also introduces her own distinctive ideas, such as constructive patterns and generalization of models for the implementation of arguments.

Conclusion

Thus, we conclude that disagreements and alternative points of view contribute to the constant development and improvement of linguistic theories. Such debates eventually lead to a deeper understanding of the implementation of the arguments and open up opportunities for further research in this area. Both C. Fillmore and B. Levin and M. Rappaport Howav have made significant contributions to understanding the surface and deep structures of language, although their approaches and terminology may differ. From the point of view of comparative analysis, although both Fillmore and Levin and Rappaport Hovav explore the relationship between the meanings of verbs and the structure of arguments, the latter's approach expands Fillmore's structure. Levin and Rappaport's inclusion of thematic grids/frameworks and their consideration of the diverse syntactic constructions in which verbs may appear provides a more complete analysis, including both surface and deep structures. Their work highlights the interaction between verb meaning, argument structure, and syntactic implementation of these structures, explaining the complex nature of linguistic organization. The approaches of Fillmore and Levin/Hovav have made significant contributions to our understanding of language.

Fillmore's work on case grammar focuses on surface structures, especially syntactic and grammatical aspects of case realization. He focuses on the comparison of deep cases and their surface expressions, considering the morphological and syntactic forms that implement these cases. The structure of Fillmore's case grammar aims to capture the relationship between the meanings of verbs and related arguments in the surface structure of sentences. On the one hand, Fillmore's approach, known as frame semantics, is based on how words and phrases evoke a mental structure or "frame" that helps us interpret meaning. This approach emphasizes the importance of context in understanding language and focuses on how words are used in certain contexts and how these contexts shape their meaning. It covers rich contextual and situational aspects of meaning.

On the other hand, Levin and Rappaport's approach, known as VerbNet, focuses specifically on verbs and their syntactic models. They classify verbs into semantic classes based on their behavior, providing a valuable resource for understanding how verbs are used in various contexts, providing a more detailed analysis of how verbs interact with their arguments and how this interaction affects the overall meaning. Levin and R. Hovav, based on Fillmore's ideas, also consider the relationship between the meanings of verbs and the structure of arguments. However, they introduce the concept of "thematic grids" or "thematic roles" to represent potential arguments related to verbs. Their analysis goes beyond the surface structures and explores how verb meanings and thematic roles interact with the syntactic and semantic properties of arguments. They consider various ways of implementing verbs in various syntactic constructions, covering both surface and deep aspects of the linguistic structure.

Further work on the legacy of Fillmore and Levin/Rappaport in the field of semantics and cases is of great interest to the linguistic community. Prospects are opening up for the advancement of lexical and grammatical analysis of language in the study of deep cases and the structure of the verb, as well as the relationship between the verb, noun, agent and patient.

The next stage of research may be aimed at developing computer models for natural language processing using programming languages and artificial intelligence. These models will allow for lexical and grammatical analysis of the language at a deeper level and identify hidden patterns in the structure of sentences. The use of computational linguistics and machine learning methods for automatic extraction and analysis of the meanings of deep cases in large text corpora contributes to the establishment of more accurate statistical patterns and relationships between verbs, arguments and deep cases. Developments in this field are also an important component of a professional approach to research. An important area of research may be the analysis of semantic relations between different parts of speech and their impact on the understanding of text by computer systems. Research in the field of semantics and cases can lead to the creation of new methods of natural language processing and improve the quality of language models for computers. Further research on the work of Fillmore and Levin/Rappaport represents an important step in the development of linguistics and computer technology.

It is also worth mentioning that the development of research in the field of cognitive linguistics and neuro-linguistics may provide new insights into how deep cases are related to cognitive processes and the representation of meaning in the brain.

However, it is important to note that further analysis and development in this area require careful comparison and synthesis of various linguistic approaches and data. This will allow you to create a more complete picture and better understand how deep cases function in different languages and communicative contexts. The study of deep cases by Fillmore and Levin/Rappaport presents significant prospects for further research in the field of lexical and grammatical analysis of language.