Do Bashal and Hepsō really mean ‘boil’? A preliminary study in the semantics of biblical Hebrew and Septuagint Greek

to re-evaluate traditional assumptions about the meanings of these words. Clarifying the meaning may illuminate some exegetical difficulties.


Introduction
Modern readers of ancient texts written in languages with no living speakers are susceptible to uncritically importing assumptions made about meaning by those who produced the lexicons and dictionaries that helped the learner in learning the language. For many lexical items, the lexicographic tradition is quite sound, yet one rarely pauses to question whether that is always the case, especially for words that are assumed to be clearly established despite the relative lack of linguistic and conceptual evidence about their meaning afforded in our texts. Presuppositions and assumptions about meaning have an impact on contemporary Bible translation and biblical exegesis, as translators and scholars often work unaware of the potential for interference in their interpretations from their 'foreign frame of reference' (Du Toit & Naudé 2005:35).
It is well known that a lexicon or dictionary is not the final answer for a word's meaning. Different dictionaries offer different explanations for terms, as Naudé and Miller-Naudé (2018:2-4) have demonstrated for the term ‫ז‬ ‫רֶ‬ ֶ ‫א‬ ['cedar'], where the dictionaries offer contradictory information. Analysis of the meaning of a word requires consideration of cognitive and cultural information, but dictionaries typically focus on linguistic information (Naudé & Miller-Naudé 2018:4).
The limitations of the available lexicons for biblical Hebrew and Greek are well known (Lee 2003;Noonan 2020:79-82). For example, the dictionaries of Hebrew and Greek tend to rely too much on material inherited from predecessors and reduce a word's meaning to a series of glosses rather than providing definitions (Lee 2004:68-70;Naudé & Miller-Naudé 2018:2-3;Noonan 2020:79-82). Despite these warnings about the limits of our lexicographic tools, biblical Hebrew or Greek is read using glosses derived from dictionaries and translations, with one rarely considering why it is assumed that one knows the meaning of any given word. Usually, coherent meaning can be The meaning of any given lexical item emerges from an analysis of its contextual usage, but with biblical languages, often a traditional gloss will be accepted as if it were the clear meaning of a lexical item. Lexicons and dictionaries rarely go all the way back to a fresh analysis of the actual usage of a lemma, so the traditional meaning is rarely reconsidered. Those learning biblical languages accept the lexicon's judgement without stopping to reflect on how the lexicon reached its conclusion. The acceptance of a traditional gloss as the meaning then gets assumed as more and more texts are read, reinforcing the assumption that the meaning is the gloss learned early on as the equivalent for that lexical item. This study focuses on the biblical Hebrew root ‫בׁשל‬ and the Greek verb ἕψω, both traditionally understood to denote the activity of preparing food by boiling in liquid. These lexemes appear in the Old Testament in some exegetically challenging texts, so clarifying the range of meaning may help to illuminate the interpretation of difficult passages. The full context of the usage of these lexical items in the Hebrew Bible and Septuagint is examined using cognitive semantics with the goal of identifying the semantic and pragmatic clues that point to the conceptual meaning evoked by these lexical items in biblical Hebrew and Greek. derived from what one reads; thus, one fails to consider the source of traditional meaning. Perhaps the tradition is sound for most terms, but it will not be known unless a term is reevaluated based on a fresh review of its usage. Lee (2017) believed that such a reconsideration of meaning is necessary.
This article is a preliminary attempt to revisit our assumptions about meaning with the terms ‫בׁשל‬ and ἕψω using cognitive linguistics, especially its attention to how encyclopaedic knowledge informs construal of a term's meaning. In the absence of native speakers of biblical Hebrew and Hellenistic Greek, how can the conceptual content that informs a lexical item's construal be determined (i.e. the encyclopaedic knowledge)? And how can it be ensured that our analysis is not being coloured by our modern encyclopaedic knowledge activated by the modern language gloss, rather than the meaning inherent in the Hebrew or Greek term?

Methodology
Cognitive linguistics provides a framework for understanding meaning as arising from actively negotiated mental representations through dynamic construal of meaning (Croft & Cruse 2004;Cruse 2011:51-68, 119-124;Langacker 2008:27-54). Meaning consists of conceptualisation and arises from the interaction between conceptual content and construal of that content (Langacker 2008:43). The conceptual categories that inform construal arise from 'ordinary human experience' (Croft & Cruse 2004:7). These categories have been given various labels such as 'base', 'cognitive model', 'domain', 'frame', 'mental spaces', 'profile', 'base', 'purport' and 'schema', among others (Croft & Cruse 2004:8, 15;Cruse 2011:119;Fillmore 2006:373;Langacker 2008:32-44). Determining the meaning of a word is a matter of construal because it involves profiling the term against the right conceptual or cognitive domain (or frame or base, etc.; this article primarily uses 'domain'). The relationship between content and construal is interactive, because lexical items work together to both imply a particular construal or evoke some content associations (Langacker 2008:43). The aspect of lexical entailment is also relevant to construal, because it involves 'logically necessary, context-independent relationships' among concepts -that is, if something is A, then it is also B (Cruse 2011:28-29). A common example is that 'it is a dog' entails 'it is an animal' but not 'it is a pet' (Cruse 2011:28). The question to be considered here is how to know what ‫בׁשל‬ and ἕψω entail. Related to entailment are the notions of constraints and default construal; constraints such as human cognition, the nature of reality, social convention and context limit the possibilities for construal (Croft & Cruse 2004:101-103). In the absence of sufficient contextual information, the default construal will be the one created by the conventional constraints (Croft & Cruse 2004:104). If ‫בׁשל‬ and ἕψω entail the idea of cooking an edible item in heated liquid in a container, then the default construal of those terms would be 'boil' in the absence of further contextual information about the process of preparation of the food. But it is possible that the construal 'boil' is activated by context, and when that contextual information that makes the action of 'boil' explicit is absent, the default construal should be 'cook'. Peters (2016) provided a valuable study of the domain of cooking terms in biblical Hebrew, grounded similarly in cognitive linguistics. His work is especially useful for taking seriously what we know about domestic life in ancient Israel and about how cooking was carried out. This background information is essential for establishing the encyclopaedic knowledge that a native speaker of Hebrew in the Iron Age would have possessed (Peters 2016:57). In surveying the verbs for cooking, he offered a detailed look at the term ‫בׁשל‬ (Peters 2016:97-105). It should be observed that his survey does not include any generic word for simply 'cook'; the closest would be the use of ‫עׂשה‬ ['make'] for preparing food.

Re-considering ‫בׁשל‬
The verb ‫בׁשל‬ occurs 28 times in the Hebrew Bible, 26 times in the piel or pual, once in the qal and once in the hiphil (reading the difficult form in Ezk 24:5 as piel). The uses in the qal (Jl 4:13) and hiphil (Gn 40:10) relate to ripening agricultural products and are thus excluded from the domain of cooking terms. The related adjective ‫ל‬ ֵ ‫ָׁש‬ ‫ּב‬ occurs twice (Ex 12:9; Nm 6:19), and the noun ‫ְלֹות‬ ׁ ‫ַּש‬ ‫ְב‬ ‫מ‬ is used once (Ezk 46:23) (Peters 2016:97-98). Based on context, the adjectival and nominal forms appear to also belong to the conceptual domain of cooking terms.
The traditional glosses for ‫בׁשל‬ in the piel or pual stems are 'boil, seethe' (BDB, s.v. ‫ל'‬ ַ ‫ָׁש‬ ‫,)'ּב‬ 'boil, cook, fry' (HALOT, s.v. ‫,)'בׁשל'‬ 'boil, cook' (DCH, s.v. ‫,)'בׁשל'‬ 'ripen, boil, mature, cook, roast' (Jastrow 1903:199) and 'kochen [cook, boil], machen [make, produce]' (Gesenius 1921:120-121), where 'kochen' glosses the uses in the qal and piel and 'machen' the one use in the hiphil (German provides an example of how a language can use one lexeme to cover both senses). The distinction among stems between agricultural and cooking domains (distinguishing the piel or pual from the other stems) may not hold in rabbinic Hebrew (see b. Sanhedrin 95b where Jastrow [1903:199] reads ‫לבשל‬ as a piel in the phrase ‫פירות‬ ‫,לבשל‬ 'to ripen the fruits', although pointing as qal seems possible). Jastrow's data for usage of the root ‫בׁשל‬ in rabbinic literature in both Hebrew and Aramaic suggests the verb does not imply anything about the mode of cooking but functions as a general word for 'cook', but a full review of the rabbinic Hebrew usage is beyond the scope of this study. Similarly, the information for ‫בׁשל‬ provided in other Aramaic reference works suggests that the semantic domains are not delimited by verbal stem in Aramaic, with usages assigned to the gloss 'cook' in the peal and to the gloss 'ripen' in the pael (see Sokoloff 1990Sokoloff :115, 2002CAL, s.v. 'bšl'). However, the focus of this study is examining the usage in biblical Hebrew, not a detailed examination of the root's usage in Aramaic either. As the Aramaic and rabbinic Hebrew data come from a later period than biblical Hebrew, it is unlikely that the data offer an independent witness of the root's range of meaning in Jewish dialects because of the potential conceptual interference of an exegetical harmonisation. Regardless, http://www.hts.org.za Open Access while the lexicons offer the glosses of 'boil' or 'cook', they do not help a reader of the Hebrew Bible to determine which sense to apply in any given passage.
Hebrew lexicography has long involved comparative data from other Semitic languages. The comparative data from the lexicons cited here and from others consulted by Peters (2016:99) show the same general options of 'cook' or 'boil' for cognates in Aramaic, Syriac, Ethiopic, Mandaic, Ugaritic and Akkadian. On the basis of comparative data alone, the weight would seem to be on the root being a general word for 'cook', although one gets the sense that the comparative data offered is incomplete. The three occurrences of ‫בׁשל‬ in the obscure commandment about a kid and its mother's milk (Ex 23:19; 24:26; Dt 14:21) are excluded from this count, because it is not completely clear that milk is being used as a cooking liquid, despite the usual translation of the phrase in question as 'you shall not boil a kid in its mother's milk'. Schorch (2010) made a convincing case that the prohibition could refer to suckling kids and should be translated 'you shall not boil a young goat which is at its mother's milk' (p. 129). If Schorch is correct, then milk in these verses is not a liquid being used for cooking.
If the debated case of ‫בׁשל‬ 'with fire' ‫ׁש(‬ ֵ ‫ּא‬ ָ ‫)ב‬ in 2 Chronicles 35:13 is also set aside, 12 cases remain where context offers no explicit indications that the specific act of cooking by boiling a liquid in a pot is in view when ‫בׁשל‬ is used (Ex 29:31; Lv 8:31; Nm 6:19; Dt 16:7; 1 Sm 2:15; 2 Sm 13:8; 1 Ki 19:21; 2 Ki 6:29; Lm 4:10; Ezk 46:23-24 [3x]). In all of these passages, cooking is clearly in view, and the item being cooked is usually meat (with the exception of 2 Sm 13:8), but reading ‫בׁשל‬ in those verses with the basic meaning of 'cook' works just as well as the more specific 'boil'. There are 10 cases where boiling seems clearly in view, three cases where it is probably in view and 12 cases where context does not provide enough information. Even if the 'boil a kid' verses are taken into account, they would not tip the balance, regardless of how one interprets the phrase.
The problem is how to account for the 10 cases where boiling is made clear from contextual information. Do they reveal information that ‫בׁשל‬ entails by default, or was it necessary to provide those details to force the construal of ‫בׁשל‬ as 'boil'? The common conclusion seems to be that it entails the presence of pot and liquid by default. For example, Peters (2016:100) stated, 'given no other context, one would assume it entails boiling or stewing something in liquid'. However, there are not enough examples to make that determination from the linguistic data alone.
The options for the meaning of the verb ‫בׁשל‬ are that it is a superordinate simply meaning 'cook', without denoting the manner of cooking, or that it is a term specifically meaning 'boil'. It is also possible that its meaning changed over time from a general word for 'cook' to a specific word for 'boil'. In addition, it could cover both options, as German's 'kochen' does, but context or cultural knowledge would be necessary for readers to make the correct construal.
Past interpreters have been drawn to the possibility that ‫בׁשל‬ does not specifically mean 'boil' as a means of reconciling the commands about how to prepare the Passover lamb. For example, McConville (1984) remarked: [T]he word bāšal in itself is so general a term for the preparation of food that the question of contradiction with the requirement of Exodus 12:9 (where the significant phrase 'in water' qualifies the verb) hardly arises. (p. 117) This conclusion is a clear oversimplification of the linguistic data, which has been shown to be evenly split. Interestingly, Peters (2016:184) essentially claimed the opposite, stating that 'there is an observable trend towards liquid cooking methods' with the use of ‫.בׁשל‬ Levinson (1997:72-73) argued that Deuteronomy has transformed the earlier instruction from Exodus 12:9 about how the Passover lamb is to be prepared: 'The roasting of the lamb is accommodated to the normal sacrificial protocol that involves boiling of those parts that the celebrant and priest share'. The interpretation of Exodus 12:9 and Deuteronomy 16:7 depends on what exactly is entailed by ‫.בׁשל‬ There are good reasons to conclude ‫בׁשל‬ means 'cook' and that the construal of the sense 'boil' must be activated by the presence of other terms. Firstly, Exodus 12:9 says to not ‫בׁשל‬ 'in water'. If the verb means 'boil' and automatically entails the presence of liquid and container (i.e. they are part of the base or conceptual domain), why do those elements ever need to be explicitly mentioned? They are not necessary for communication if they are entailed by the base. Secondly, the Talmud is silent on these apparently conflicting instructions about the method of preparation where the rabbis discuss Exodus 12:9 (b. Pesahim 40b-41a; 74a), and Mekhilta d'Rabbi Ishmael claimed that ‫בׁשל‬ means 'cook' (Mekhilta Bo, vi). Thirdly, the medieval commentators such as Rashi and Ibn Ezra insist (like Mekhilta) that it can mean 'cook' (Peters 2016:178-179). Fourthly, the Hebrew usage allows the superordinate meaning in most cases.
However, there are also good reasons to conclude that the verb means 'boil'. Firstly, the long history of attempts to reconcile the commands about Passover preparation (beginning with 2 Chr 35:13) suggests an awareness that the verb typically meant 'boil'. Levinson (1997:70-73) considered Mekhilta to also reflect a 'harmonistic solution' to the commands rather than an independent witness of the meaning 'cook' (which in turn suggests the possibility that the meaning 'cook' in rabbinic Hebrew and Jewish Aramaic could have been influenced by this harmonisation). Secondly, the use of the term in contexts with other cooking verbs such as 'bake' would be unusual if it were a superordinate for 'cook' because there is no reason to contrast a general word such as 'cook' with a specific word such as 'bake'. Thirdly, it is likely that Hebrew would have had a specific term for 'boil' given the importance of boiling as a cooking method; if it is not ‫,בׁשל‬ then such a term appears to be lacking (Peters 2016:184). In fact, the dominance of boiling as a cooking method suggests that it could be the default construal for ‫;בׁשל‬ lacking other contextual clues, the reader or hearer would assume the word meant 'boil'.
This survey of contextual usage of ‫בׁשל‬ was intended to step back and re-evaluate our traditional assumption about what the root meant. Based on the cultural and conceptual information related to cooking terms, it seems that the assumption it refers to 'boiling' most of the time is supported; the biblical writers likely then had other reasons for mentioning explicit cooking elements present at the scene, such as liquids and pots, if the verb alone activates a default construal of 'boil'. However, in context, perhaps it is sometimes simply a generic word for cooking, similar to how German's 'kochen' means 'cook' and 'boil'. The fact that comparative data for other Semitic languages cover both senses continues to hold out that possibility for biblical Hebrew, and our lexicographical tradition indicates that with their suggested glosses.

The meaning of ἕψω
Now let us briefly turn to the Septuagint and its usage of ἕψω. Most occurrences of ‫בׁשל‬ in the Hebrew Bible are translated with ἕψω in the Septuagint (exceptions include 1 Sm 2:15; Ex 34:26; and 1 Chr 35:13). The verb also translates ‫זיד‬ in Genesis 25:29 where Jacob is preparing a stew. As with ‫,בׁשל‬ the traditional glosses are 'boil, seethe' (LSJ, s.v. 'ἕψω'), 'cook' (Sophocles 1900, s.v. 'ἕψω') and 'to make [something] cook, boil' (Montanari, Goh & Schroeder 2015, s.v. 'ἕψω'). Unlike ‫,בׁשל‬ there is a great deal of linguistic evidence for ἕψω outside of its use in the Septuagint. The Greek lexicons list a decent, representative cross-section of this evidence for ἕψω. Doing an adequate analysis of the usage for other lexemes might require consulting digital tools such as the Thesaurus Linguae Graecae, 1 which provides hundreds of examples for ἕψω, even in the abridged version. This body of evidence may help shed light on ‫בׁשל‬ inasmuch as the Septuagint translators considered it the best equivalent for the Hebrew verb. The verb ἕψω appears with much fuller context, describing cooking elements and methods in Herodotus (e.g. 1.48.2; 4.61) and other Greek writers; this context demonstrates much more clearly that ἕψω generally means to cook by boiling. Its usage in the Septuagint of Genesis 25:29 reinforces the conclusion as the context is one of the two clear examples in the Old Testament where stew is being prepared (the other is 2 Ki 4:38). Finally, the contrasting usage of ὀπτάω and ἕψω in the Septuagint of 2 Chronicles 35:13 for the two occurrences of ‫בׁשל‬ suggests that the translator understood the nuance of meaning and was, like many interpreters since, harmonising the competing statements on the method of cooking the Passover from Exodus 12:9 and Deuteronomy 16:7. The translator realised that ἕψω was not an appropriate equivalent for the sense intended for the first occurrence of ‫בׁשל‬ in 2 Chronicles 35:13.
With ancient Greek, the vast corpus of literature often means there are many more occurrences of a word to analyse, but those occurrences help in reconstructing the encyclopaedic knowledge necessary to define the frame, base or cognitive domain that informs the construal of a word's meaning.

Conclusion
Readers are still susceptible to making assumptions about a word's meaning and uncritically accepting the glosses from a lexicon. When one reads biblical passages, it is very difficult to not automatically assume the gloss one has internalised as the meaning. The proper interpretation of a passage may hinge on the meaning of a specific word, but interpreters often assert the meaning of a term without explaining why the term carries that meaning or indicating how they have come to the conclusion that it has that meaning. Turning to new linguistic approaches also has its challenges. Determining what the encyclopaedic knowledge informing a term's construal was can be difficult, absent the availability of volumes of usage in extra biblical sources or clear archaeological data.
With this preliminary study of ‫בׁשל‬ and ἕψω, it appears that the lexicographic tradition is on solid ground with the meaning 'boil'. With the Hebrew term, it is likely the default construal of the verb. This conclusion seems to be the only way to explain the harmonising tendency evident in 2 Chronicles 35:13. With the Greek term, the contextual usage fills out the encyclopaedic knowledge for the term to such an extent that the meaning seems well supported. In turn, the fact that the Septuagint translators felt it was the best equivalent for the Hebrew verb reveals they had the same understanding of 1.See http://stephanus.tlg.uci.edu/index.php