How to Analyse the Meaning of “Too” and “Also”

Abstract

The words “too” and “also” seem interchangeable, yet their meanings differ in subtle ways. This essay analyses and compares their use, drawing on semantics and pragmatics, conceptual and procedural meaning, implicature and relevance theory to show how such differences in meaning can be studied.

Introduction

Language is used for communication, but the interpretation of an utterance in communication is only partially determined by the meaning of the words and constructions used by the speaker. An utterance can have multiple interpretations that arise from linguistic expression and their meanings, this essay will show that interpretations are heavily dependent on cognitive principles, such as existing knowledge between speaker and listener, cultural backgrounds, emotional state and context. In this essay, the words “too” and “also” will be analysed and compared. The similarities, overlap, and differences in use will be discussed and evaluated. This essay will also demonstrate how one goes about analysing those differences and similarities in meaning by clarifying: semantics and pragmatics, conceptual and procedural meaning, implicatures, relevance, and mental processes to use as tools to analyse “too” and “also”.

Semantics and Pragmatics

Communication can be viewed from different perspectives as it does not depend solely on recognising the meaning of words in an utterance, it depends on the context and intentions of what the speaker is actually intending to say. Semantics is the study of the meaning of words, phrases and sentences. The primary focus is on what the words conventionally mean, rather than speaker’s intentions or what someone means on a particular occasion. Semantics is concerned with general meaning and has no room for subjective meaning (Yule, 2010). Pragmatics is the study of the intended meanings that are not conveyed by speech itself, but humans recognise those meanings gained through experiences from their developmental years to present time. Humans have the ability to predict each others’ behaviors and understand why others behave in the way they do. Humans also have the ability to empathise with emotions of others and understand that people may see things from a different perspective. To understand these invisible meanings, humans must rely on shared assumptions, context and expectations when communicating. Utterances occur spontaneously and interpretations happens so fast that people are unware how they have understood the message and give a response in milliseconds (Horn and Ward, 2005). The mind decodes what has been said and uses background knowledge and context to work out the intended speaker meaning. Traditionally, pragmatics highlights the roles of intention, rationality, mutual knowledge, relevance, and commitment in forming and executing communicative acts (Horn and Ward, 2005), which means pragmatics is an inference-based discipline. However, utterances are hard to define, because the meaning of a linguistic expression is conceptual when it encodes a concept and information about intentions (Wilson and Sperber, 2012). Conceptual encoding is when a word encodes a concept which acts as the basic logical form of a sentence in which that word occurs, and procedural encoding is where a word or expression encodes information to guide the hearer during the inferential phase to a specific sentence comprehension (Wilson and Wharton, 2006). For example, words that are homonyms are written the same way and encode two different concepts (e.g. bat). A bat can be an object to hit a ball or the mammal that flies in the night, the context of the sentence makes it clear to people which “bat” is meant without even considering the other option. According to relevance theory, pragmatic processes are inferential (Blakemore, 2004). Inference is additional information used by the listener to create a connection between what was said and what must be meant. (1a)Can I borrow your Chomsky? (1b)Sure, it’s on the top shelf. When people refer to something, as the example given in (1a), it is assumed that the listener can recognise which referent is intended by the speaker, due to their world knowledge (i.e. if X is the writer of the book, then X can be used to identify a copy of a book by that writer). The listener knows through inference and context that it is not meant as literally borrowing “Chomsky”, rather the copy of Chomsky's book. People also include memories of particular occasions and about particular individuals, general cultural assumptions, religion, science, or the emotional state both people are in. People select the right assumptions in order to interpret the intended meaning (Blakemore, 1992). Utterances seem to be linguistic messages which are borne out of assumptions of what the recipient knows. What a person assumes is described as presupposition (Yule, 2010). The presupposition in the example above (1a) is: you have a copy of Chomsky’s book. Utterances carry meanings of actions such as: requesting, commanding, questioning, informing or promising. For example: (2) I’ll give you your Chomsky back tomorrow morning (2) is an utterance performing the speech act of promising. The framework of speech act theory illustrates a distinction between semantics and pragmatics of describing and indicating (Blakemore, 2004). Blakemore (2004) illustrates that this distinction is truth-conditional and non-truth-conditional lies within discourse connectives. After Grice’s theory in 1975, the truth-conditional semantic theorists had to acknowledge that some linguistic items did not seem to play a role in the purpose of propositional content, their input seemed to have more of a pragmatic influence similar to presuppositions or as a predictable implicature. For example, “therefore” links two propositions and presupposes, or implicates, that the second proposition introduces a consequence for the first proposition (i.e. cause and effect). In contrary to “but”, which seems to depend on the truth of two propositions presented as “and” does, but adds something else which makes it non-truth-conditional and includes pragmatic competence (Blakemore, 2004). Relevance Theory (RT) is a combination of Gricean intention-based pragmatics and cognitive science to provide a cognitive-inferential pragmatic framework (Wharton, 2009). Ostensive behaviour is when the communicator provides evidence of an intention to communicate something other than semantic meaning. Wharton (2009) argues that utterance interpretation is a two-phase process: decoding of the grammar which is the inferential process to conform standard communication, and the ostensive behaviour which will create the assumption that there must be a reason why something is communicated. For example:

Adam is late for a meeting. He knows he is late, but his boss points at the clock or frown towards the clock and utter the words that he is late. The reason that Adam’s boss is communicating the message (which Adam already knows) is because he wants to let Adam know that there will be consequences for his actions.

An input is relevant in context when it interacts with that context to produce cognitive effects. Cognitive effects strengthen a contextual assumption or contradict and eliminate the contextual assumption, or combine textual assumption to produce a contextual implication or prediction (examples will be provided in the next section). Intuitively, relevance comes from thought, memories and conclusions of inferences. An utterance becomes relevant to an individual when it connects with the background information that is available to draw a conclusion (Sperber and Wilson, 2012).

Implicature

Blakemore (2004) shows that discourse markers are better understood when placing restraints on the inferential phase of interpretation, by providing instructions to the recipient on how to process the proposition for relevance. The recipient is then guided to the obvious choice of contextual assumptions where they understand the intended interpretation. Discourse markers are used to point out the transition of inference, and contribute to utterance interpretation by creating inferential relations between propositions in conversation. This is done to strengthen contextual assumptions and highlight implications. Blakemore (1992) uses a procedural approach to analyse discourse connectives in which she accounts for particles, such as “also” and “too”, interacting with focus to be syntactic structures with cleft constructions, and that intonation interacts with both conceptual and procedural meaning. Iten (2005:23) refers to these adverbs as ‘focus particles’, which do not contribute to the truth-conditions of a sentence. I would prefer to call these items “additive particles”, because the particles “too” and “also” add information to make the implicature more obvious. For example: (3a) Are you coming out tonight? (3b) I’ve got a lot of reading to do. (3c) I’ve got a lot of reading to do and I also need to finish my essay. (3d) I’ve got a lot of reading to do and I need to finish my essay too. (3e) No, I’m not coming out tonight. The conclusion (3e) is derived from the two propositions together, it eliminates assumptions that could arise by having the first proposition on its own. When the speaker puts the stress on a lot or reading (3b) it does not eliminate the possibility of “going out” (e.g. I’ve got a lot of reading to do, but I could use some distraction). The additional information in (3c) and (3d) strengthens the implied message of the speaker and eliminates the possibility of “going out”. It can be said that “too” and “also” make the intentions of the speaker more explicit. However, in this example, the use of these “additive particles” also indicate that meaning is derived from the first proposition, and that the second proposition cannot exist in isolation (which includes “too” or “also”). Reading the previous sentence, it is clear that “also” (underlined) can be used to add information to an example used, which “too” cannot perform.

(4a) Maks ate cookies . (4b) Sermin ate cookies . (4c) Sermin ate cookies too. (4d) Sermin also ate cookies. To achieve relevance in a conversation, and be explicit with the information the speaker wants to provide, to add information to (4a) the obvious response would be (4c) or (4d). (4b) suggests parallelism of the syntax and encourages the recipient to derive parallel implicatures (Blakemore, 1992). These examples provide evidence that “too” and “also” are non-truth-conditionals, as they add meaning, whereas “and” would only connect (4a) and (4b). (5a) Maks also ate cake. (5b) Maks ate cake too. In (5a) and (5b) it suggests that Maks ate something else besides the cake or there might be someone else involved. The variable of someone else is given in the examples of (4c) and (4d). There might be a difference between these utterances, one could entail something else has been eaten by someone (6a), and the other that someone else has eaten multiple things (6b). The relevance of the utterance will depend on which entailment will be taken as background information. (6a) Maks ate cookies. He also ate cake. (6b) Sermin ate cookies. Maks also ate cake. (6c) Sermin ate cookies. Maks ate cake too. (6d) Sermin and Maks ate cookies. He also ate cake. (6e) Sermin and Maks ate cookies. He ate cake too. The recipient needs the context of (6d) to understand (6b), “also” creates the assumption that Maks ate cookies and cake, whereas “too” cannot perform that act (6c). When the speaker utters (6c) the recipient would intuitively assume this to be a slip of the tongue, because cookie encodes the concept of “cookie” and not cake. To make this more obvious, when the context is clear (6d) suggests that two people ate something and a male ate another something, that male could be Maks, it suggests “a male” ate cookies and cake. In (6e) “too” makes it explicit that “he” must be Maks, and if it is not Maks, the sentence would not make sense as in (6c).

The meaning of “too” seems to change when grammatical structure changes as shown in (7c). (7a) Lucy is tired too. (7b) Lucy is also tired. (7c) Lucy is too tired.

Examples (7a) and (7b) show similar assumptions to the previous examples. The entailment in (7c) intensifies the meaning of “tired” and entails that someone is very something. (7c) seems to express explicitly that Lucy is extremely tired and eliminates the assumptions of being “a little tired” and shows that “too” acts here to emphasise. Thus, the mind evokes other assumptions, it creates questions of why Lucy could be tired, because it implies Lucy is too tired to do something. (8a) Laura is Dutch. (8b) She is, too. (8c) (pointing at someone) She is too. (8d) She is, also (8e) Paula is too

The construction in (8b) is intended as a confirmation of the claim in (8a) which (8d) cannot seem to perform. However, (8c) and (8e) are claims that another person besides the person in (8a) is also something, and again (8d) cannot perform that claim either.

Mental Processes

People that learn a second language rely on their first language for background knowledge. A similar process happens when people think in one language and have to retrieve knowledge and representations to make sense of the world (Saville-Troike, 2012). Mental processes help people to make decisions and those processes are linked to language. In order to understand example (10), a set of conceptual representations, linked by inferential computations by two people, will be presented in order to understand their decisions and language processing at the time of conversation (which will be exemplified in table 1). (10) (Sermin and Maks talking over lunch. Their mental processes of what happened before lunch is available in table 1) (10a) Sermin: I bet you cannot lift me? (10b) Maks: I could lift you, if you hadn’t eaten too much (rolling his eyes). (10c) You have eaten too much. (10d) If you have eaten too much, I cannot lift you. (10e) Did Sermin get lifted by Maks? (10f) Sermin did not get lifted by Maks.

Sermin on Monday morning Maks on Monday at the cafeteria Premises: Sermin is trying to get ready to go to a lecture, when her pants won’t fit her (visual perception) I ate too much over Christmas (language processing) People that eat too much, get fat(memory) Premises: Maks sees Sermin struggling at the cafeteria to pick her lunch. She gets a fruit salad (visual perception). She is trying to lose weight again (language processing) Sermin gained 10Kg since she moved here (memory) People try to lose weight in January (added memory) Conclusion: If I eat too much, I’ll get fat Added premise: Exercise and healthy food will get rid of the fat Conclusion: If Sermin picked a salad she wants to lose weight Added premise: People that eat healthy lose weight Conclusion: If I don’t exercise and eat healthy, I’ll get fat Added premise: I don’t want to be fat Conclusion: If she picks the other options in the cafeteria, it might not help her to lose weight Added premise: She needs the help Conclusion: I should start exercising and eat healthy from now on Conclusion: I should not try to change her mind Table 1. Conceptual representations and inferential computations (template adapted from Wharton 2016) Maks’s utterance (10b) will be true if and only if he lifted Sermin before she ate too much and was successful at lifting her. However, it shows an oddness between Maks not being able to lift Sermin and Sermin eating too much (when she only had a fruit-salad). Maks could have the intention of being funny, because Sermin only had a fruit-salad for lunch. As mentioned before, intonation and stress are also elements that need to be considered to comprehend the meaning of an utterance. Maks has an unusual rise and fall in his voice and is signalling with his eyes. However, Sermin will think that Maks is actually implying that she is too heavy, due to her thought process of that morning, and pays no attention to Maks’ overt information transmission. In other words, there is an assumption: If x is not able to lift y, then y is too heavy for x to lift. According to the code model, Maks encodes the intended message into a signal, which is decoded by Sermin using the identical copy of the code. According to the inferential model, Maks provides evidence of his intention to convey a certain meaning, which is inferred by Sermin on the basis of the evidence that is provided in her current state of mind. The utterance is a linguistically coded piece of evidence. Thus, verbal comprehension involves an element of decoding which can be misinterpreted, because the utterance is relevant to Sermin when it is processed in a context of available assumptions in the mind. This is also known as contextual implication, which is a conclusion deducible from the utterance and the context together (Sperber and Wilson, 2012). Linguistic decoding is seen as connecting a construction of series of representations (see table 1) by a variety of computations (Wharton, 2016). The analysis of visual perception is seen as a linking construction of a series of abstract visual representations, linked by a variety of computations. People integrate information from different sources, the example in table 1 shows information from visual perception, linguistic processing and memory combined. Blakemore would argue that Maks’s conclusion is derived from the visual which can be considered as contextual implication which was strengthed by existing assumptions. Maks Visual perception linguistic process memory conclusion (Relevance Theory) (sees what Sermin is doing) contextual implicature existing assumptions elimination of existing assumptions conclusion (Contextual Implication) Wharton (2016) argues that conceptual representations have a variety of properties that distinguish them from other types of representations: “Conceptual representations give information from all senses and can be integrated and compared; Their constituents are concepts rather than syntactic categories of phonetic features; They are more logical than linguistic acoustic properties. They can be true or false, can imply and contradict each other; They act as the input to logical rather than syntactic or phonetic computations.” (Wharton 2016:2) Relevance theory proposes that both assumptions and conclusions should be regarded as implicatures of the utterance, and that humans will automatically maximise relevance due to the evolution of the cognitive systems (Sperber and Wilson, 2012). (10d) is an implicature assumption and (10f) is an implicated conclusion. The utterance produced by Maks indicates that he expected Sermin to assume (10f), and of course, Sermin was already considering the existing assumption of (10f). This means that Maks’ utterance would be to confirm her belief is the truth (Blakemore, 1992). Blakemore (1992) would argue that the assumption Sermin has interpreted, of Mak’s utterance, is not retrieved from memory or deduced from existing assumptions, whereas it is obvious that both ideas work in parallel in this particular conversation. In this scenario, “too” is used as an excessive, which is a role that “also” cannot perform. The example above does raise the question of how non-truth-conditional “too” really is. Depending on context, homonyms encode a set of possible meanings. These possible meanings will evoke an image in the listener’s mind. It has been demonstrated above, that “too” works in a homonymous way, without having an image – dependent on the grammatical structure of the sentence.

Conclusion

This essay has provided background knowledge of semantics and pragmatics, conceptual and procedural meaning, presuppositions, assumptions, implicatures, relevance, and mental processes in order to answer the question posed by giving examples that show similarities and differences in “too” and “also”. At first glance, it appears that “too” and “also” are identical as they both add information and both can be used to eliminate assumptions and contradictions. Both “too” and “also” achieve this by making the intention of the speaker more explicit (see example 3). Examples (4) and (5), however, show a different usage pattern. These examples show the importance of context and relevance to distinguish speaker meaning for both particles. In example (6), “also” seem to be able to imply other possible assumptions, in which “too” cannot seem to entail the same way, because “too” seems to depend more on context and the first proposition. Example (8) shows that “too” is able to show confirmation where “also” cannot. After the explanation of example (3), “also” showed that it can add information or context to a previous example or sentence where “too” cannot. “Too” has been labelled as a non-truth-conditional, but the final example (10) given in this essay casts doubt on this – when “too” is used as an intensifier. This example shows that “too” can be homonymous and used as a truth-conditional (intensifier) and non-truth-conditional (additive particle). (Wordcount: 3086)

Refrences

Blakemore, D. (1992) Understanding Utterances: An Introduction to Pragmatics, Oxford: Blackwell Publishers. Blakemore, D. (2004) Relevance and linguistic meaning: The semantics and pragmatics of discourse markers, Cambridge: Cambridge University Press. Iten, C. (2005) Linguistic meaning, truth conditions and relevance: The case of concessives in Breheny, R. and Bach, K. (eds), New York: Palgrave Macmillan. Saville-Troike, M. (2012) Introducing Second language acquisition. Muriel Saville-Troike (2nd edition), Cambridge: Cambridge University Press Wharton, T. (2016) 'Lecture 4: Conceptual and Procedural Information', Lecture, (4) Wharton, T. (2009) Pragmatics and non-verbal communication, Cambridge, UK: Cambridge University Press. Wilson, D. (1975) Presuppositions and non-truth-conditional semantics, London: Academic Press. Wilson, D. and Sperber, D. (2012) Meaning and relevance, Cambridge: Cambridge University Press. Wilson, D. and Wharton, T. (2006) 'Relevance and prosody', Journal of Pragmatics, 38 (10): 1559–1579. Yule, G. (2010) The study of language (3rd edition), Cambridge: Cambridge University Press.

← Back to Papers