The need for textual evidence in reconstructing linguistic pictures of conceptual categories

The study responds to Jerzy Bartmiński’s call for using various kinds of data in doing ethnolinguistic research. The authors, thus, champion (i) the necessity of including textual data in reconstructing a linguistic worldview as well as (ii) a culturally-oriented non-systemic lexical semantics. That ethnolinguistic reconstruction cannot be based merely on dictionary-like information is shown by a critical assessment of purely systemic analyses of hussy, pheasant, and maid. It is concluded that textual evidence should be extensively explored and systematically used in cognitive ethnolinguistics as primary, rather than as supplementary data.


Introduction: aims and issues
There are two main goals behind the present contribution. One is to foster the conviction, if not justify the necessity, of including textual data in ethnolinguistic research, especially in relation to the task of reconstructing the Linguistic Worldview (henceforth: LWV), as understood and practised by Bartmiński and his collaborators within the Lublin Ethnolinguistic School. As noticed by Głaz, Danaher and Łozowski (2013: 17-18;henceforth: the Versita volume), Bartmiński makes use of four kinds of data: the language system, texts, questionnaires, and "co-linguistic" data (customs and social practices, either accompanied or not by the corresponding linguistic expressions). Yet, the Versita volume, one of the latest comprehensive accounts of LWV, shows that the LWV-oriented researchers tend to be selective in their choice of the database, with probably just two contributions (by Gicala and by Vergala, that is) truly appreciating texts, rather than the system. Similarly, in Bielińska-Gardziel et al. (2014), another recent collection of LWV accounts, dictionary-based analyses clearly outnumber dictionary-cum-text ones, let alone purely textual investigations. Moreover, most recently, Niebrzegowska-Bartmińska (2015: 37), has admitted that "the Polish researchers [. . . ] assume that the base of the profiling process includes the characteristics derived from all or only some of the kinds of data" (emphasis added). As we understand, then, textual evidence may, but does not have to, be part of the methodological credo in reconstructing LWV. If so, texts still need to be shown to be of primary importance and direct applicability on the grounds of reconstructing conceptual categories.
The other of our aims is to champion a certain kind of lexical semantics, or better: a particular way of studying meaning (and changes in meaning), that the LWV-oriented research seems to favour. Namely, as soon as we are prepared to go beyond and above systemic evidence, there is a problem of relating conventions of word use with broad socio-cultural values. In other words, relating changes in the use of a given word to changes in mentality and social structures means drawing parallels across all-embracing and abstract entities, which may produce overgeneralizations as much as oversimplifications. This makes some believe that "changes in our world are neither necessary nor sufficient to bring about changes in our language" (Keller 1994: 5). For example, according to Keller, the reason why so-called women-terms tend to be pejorative has nothing to do with the diminished status of women due to industrialisation or other social developments and causes. Instead, he argues (p. 76-78) that because of the "knightly" tradition of deference attitude towards women, speakers will tend to be exceptionally careful and formal rather than taking a risk of being too little formal or not formal enough, so that formerly positive terms are always in danger of becoming just neutral, and neutral terms of becoming pejorative ones. 1 Still, our world is a construct of our cognitive and mental experience -being "knightly", being careful, being afraid of not being formal enough etc. are all experiential considerations that find their way, deliberately or not, into language use. It is our conceptualization of the world, and not the world as such, that happens to be represented in language.
How do our aims relate specifically to the LWV research? The presentday state of the LWV research seems to be sensitive to two inherently related problems. One is ontological in nature and has to do with defining the very object of examination, that is what it is that one ultimately studies in the LWV paradigm. This problem has recently been pushed significantly further within the LWV circle in Etnolingwistyka 26 (2014) and Bielińska-Gardziel et al. (2014), with a number of researchers having offered very interesting prospects of research on the ontological plane. The predominating opinion seems to be that the object of LWV investigations is extra-and super-linguistic mental constructs called koncepty (in Polish) that are not to be confused with either lexemes, meanings, concepts, ideas, or stereotypes, though the latter are probably their closest ontological analogues (cf. Bartmiński and Chlebda 2013: 71).
The other issue is epistemological and, as already mentioned, relates to the question of the data used for the LWV reconstruction purposes. The LWV researchers typically identify their material basis with the SAT strategy: the linguistic system, the questionnaire, and the text 2 , with a generous allowance of "co-linguistic" data as the fourth kind of evidence (see above). This invites questions of what and how each of the sources contributes to the overall reconstructed picture of LWV.
Naturally, these two issues are closely interrelated: the choice of the object of research affects the range of the data one needs to examine in order to approach the object, and the selection of data determines the object of research. Yet, for Bartmiński, this two-way traffic reduces to the former: it is the assumed definition of LWV that imposes the quality and the quantity of empirical evidence. This is perhaps why Bartmiński usually precedes how he means to study LWV with how he understands LWV itself : "the way we reconstruct LWV must correspond to the way we understand its existence (its ontology). If we assume that LWV is present in natural language [. . . ] as 'naïve' [. . . ], we should try to identify it in an every-day variation of the national standard" (Bartmiński 2014: 283;translation ours). 3 In this article, we explore the second of the two issues mentioned, that is the material sources of the data that may be used for the purposes of reconstructing a linguistic picture of conceptual categories, with special interest in appreciating textual data. However, instead of embarking on exploring any of the LWV research pieces, we plan to do that by a critical analysis of some independent material examined by an independent group of researchers. Specifically, we will present the diachronic semantic programme (hence: DSP) as practiced by Kleparski and his Rzeszów-based research circle. 4 In this way, we hope not only to assess selected dictionary-based analyses of DSP and their systemic orientation, but also contribute to the on-going discussion within the LWV circle on the use of different databases.
The main reasons why the DSP research seems to be relevant in the context of LWV include the following: both, DSP and LWV, aim at reconstructing a linguistic picture of selected conceptual categories, both recognize the expressive function of language 5 , both assume an axiological and a cognitively-oriented approach to their subject-matter, with grounding the language-reflected situations in the extra-linguistic reality, and both make extensive use of synchronic and diachronic perspectives. However, there is one crucial divergence: the two differ rather radically, if not drastically, in his understanding of the ontological status of LWV, and not the other way round. In this sense, Bartmiński (2009: 23-35  5 However, in DSP, the culture-expressing function of language happens to be closely related to a more controversial claim of the culture-conditioning function of language. Kochman-Haładyj and Kleparski (2011: 39; emphasis added) make this point clear: "Another function of human language, except communication, is to express shared assumptions and transmit implicit values and behavioural models to those who use it. Hence, as a powerful conceptual force, language is a transmitter of society's deep biases and provides a means of conditioning our thoughts." Evidently, language not only expresses and transmits cultural values, but it also conditions them. The DSP researchers assume, then, an uneasy blend of a cognitive linguistic claim that language is a conceptual tool and a structurally-and generatively-oriented claim that language is a conceptual force. Though this dialectics (that language can be as much of a creation as of a creator of social life, or as much of a mirror as of a perpetuator of cultural attitudes) can also be found in the LWV research, the latest developments within the LWV circle seem to point to a growing, if not determining, role of cultural parameters in shaping linguistic expressions, for which see Łozowski (2014a and 2014b). the choice of the data they use for reconstruction purposes, to which we proceed now.

Beyond and above systemic data
While Bartmiński's LWV makes use of the SAT strategy, Kleparski's DSP relies exclusively on systemic data. The methodological clash is, then, evident. Even if we allow for the fact that DSP consists in detecting the paths and causes of lexical semantic changes over relatively long stretches of time, which makes the use of the questionnaire either impossible or irrelevant, there is still a question of the textual source of information on these changes. The lack of the latter is truly striking.
How do the DSP researchers justify their limited use of systemic sources of linguistic data? They seem to assume that reading lexicographic entries "provides [one] with direct access to [a given] culture and society" (Włodarczyk-Stachurska 2011: 8), and that this access is informative enough to reconstruct the values, symbols, attitudes and mentalities in this given culture and society. In other words, the lexicographic picture of birds, women, clothes etc is, in fact, supposed to be unequivocal with the linguistic picture of birds, women, clothes etc, and the dictionary definition of cat, lady, or stone is taken to be tantamount to the conceptualization of a cat, a lady, and a stone. This is highly controversial theoretically and questionable methodologically, even if supplemented with frequent recourses to encyclopedias as a possible check-up on the generalizations being offered.
Naturally, the DSP researchers are aware of the limitations of their approach, which is, simply speaking, that the dictionary, their ultimate and exclusive point of reference, may lack substantial elements of how humans conceptualize the world and, for that reason, it needs to be complemented with other non-systemic sources of conceptual information. As to the latter, Kochman-Haładyj and Kleparski (2011) may well pronounce "the need for more data-oriented studies" (p. 17), or Włodarczyk-Stachurska (2011) may mention questionnaire-based research or observation-based research as examples of "the main paths in the research of dictionary use" (p. 9), but they would anyway derive their data from reference works like dictionaries and encyclopedias without consulting a single historical text or a single human informant.
At best, the DSP researchers admit to selected inadequacies of the lexicographic sources they rely on in their reconstructions. For example, Włodarczyk-Stachurska (2011) complains about the lack of extra-linguistic considerations in how suburb is defined in the dictionaries she has examined, claiming that no definition of suburb can do without a "comprehensive, highly informative cultural note" and concluding that "it seems obvious enough that most of the vocabulary is culture-specific" (p. 59). This is why lavatory and toilet (p. 125) may indicate the same real-world object, but they relate to two different conceptualizations of that object, of the high class and low class, respectively. Similarly, that icebox and refrigerator as well as the wireless and radio (ibid.) point to the same devices is only secondary in importance in relation to the fact that these two pairs of words reflect the age distinction and the dialect (British vs. American English) differentiation.

Hussy
But let us take a more detailed look at one of the lexemes Włodarczyk-Stachurska (2011: 167-169) examines -hussy as a phonetic reduction of the historically earlier housewife. The semantic history of hussy depicted in the OED, the primary source of systemic information for English, shows that initially hussy was employed in the evaluatively positive sense 'the mistress of a household'. Indeed, the OED evidences this with the quotation dated back to 1530 which puts hussy on a par with master and lady and sets in contrast to servant. Later, the word was associated with behaviourally negative axiological elements denoting 'a rustic, rude, opprobrious, or playfully rude mode of addressing a woman'. In the middle of the 17 th century, the word started to denote 'a strong, country woman, a female of the lower orders' and -with the progression of pejoration -'a woman of low or improper behaviour, or light or worthless character', as well as 'an ill-behaved, pert, or mischievous girl'. At this point, Włodarczyk-Stachurska complements her OED-derived description with several quotations from other lexicographic works, mostly dictionaries, all to the effect that the present-day English stylistic labelling of the word should be far more negative than <old-fashioned>, <disapproving>, or <humorous>, which is how it is usually presented in English-as-a-foreign-language (EFL) dictionaries. The reason for that is that hussy implies today -depending on the lexicographic source -'a woman or girl who is sexually immoral', 'a woman who habitually has casual sex', 'a saucy or flippant girl', 'a trumpet, trollop', 'a lewd or brazen woman', 'a saucy or mischievous girl', 'an ill-balanced girl', a 'jade', 'prostitute', or 'minx'. Now, suppose we want to make hussy, and its semantic history, part of the linguistic worldview of the English woman in general and of the English housewife in specific. We would, then, have to relate the detected changes in English to the corresponding and underlying changes in the English mentality, transformations in the English society, events in the English history, landmarks in the English culture etc. in order to identify the reasons why hussy, having been thriving in the positive sense for at least a century, started gaining its pejorative overtones, overtly negative meanings included. It is not that Włodarczyk-Stachurska cannot offer any answer, but her explanation (after Mills 1989) is again a forced dictionary-based systemic generalization that lacks specific textual evidence: Pejoration accompanied the diminishing status of a housewife as political matters and affairs passed from individual family household to centralized governments of nation states and, later, as paid productive labour passed out of the household and into the factory.
So, as we understand, as long as hussy meant 'a female manager of a household', its positive sense was guaranteed and remained parallel to that of master and lady, but once the family had stopped to be the centre of the household universe, hussy lost any sensibly positive point of reference and, thus, deteriorated.
This poses a number of uneasy questions: why should the English family perform its traditional decision-making function under the Protestant Tudors and start losing it only under the Catholic Stuarts? Why should the managerial significance of hussy start diminishing precisely under the reigns of probably the two most powerful female managers in the English history, Elizabeth I and Mary I Stuart? Why should we relate the axiological downfall of hussy to the development of industry, if the first true English factories can be dated back to the latter half of the 18 th century, which is more than a century after hussy recorded its first evaluatively negative reference to supposedly immoral Alice Pierce (or: Alice Perrers) as a concubine to the 14 th century King Edward III? Why should we assume the working class background as a motivating factor here if Alice Pierce had as little to do with that as the 14 th century did with the industrial revolution or the rise of trade unions? 6 While these doubts severely question the logic in Mill's (1989) hussy entry, here are some more issues that lay bare the inadequacies of the dictionary as an exclusive data provider and, for that reason, need further research and thorough explaining. First, hussy records its first literary attestation in Edinburgh (see hussy in the OED), which presents it as a Scottish English innovation that would gradually spread into the southern varieties of English as well. Possible, but not likely when the power struggle between the English Tudors (Henry VIII and Elizabeth I) and the Scottish Stuarts (Jacob V and Mary I Stewart) is reaching its peak, with the ultimate England-favouring resolution in the form of the personal union of England and Scotland in 1603. This meteoric rise of hussy would be as unexpected as the promotion of lad 'boy', lassie 'girl', or any other Scottish English expression.
Second, hussy shows its negative connotations as late as the mid-17 th century (1650), while the first pejorative attestations of the related baseform housewife go back in time by at least a century (1546). This not only makes Mill's industrial motivation even more feeble, but also means that the axiological histories of housewife and hussy are not independent from each other. Indeed, despite a significant 3 century difference in origin (housewife appears in the 13 th century, while hussy as late as the 16 th century) as well as a century difference in developing negative associations by the two (17 th c. versus 16 th c.), both words record the new sense of 'a case for tailoring tools (needles, thread, scissors)' in the middle of the 18 th c. This can only mean that the positive household-related dimension of hussy still lingers on at the time of first factories and of paid productive labour.
All in all, it may well be that what we have at work behind the hussy semantic history is one and the same experiential conceptualization of a wellorganized female decision-maker that reaches her objectives with whatever means available, be it thriftiness and wise economy, sexual attractiveness and seductive powers, or, simply, exploiting men's wallets. This can be axiologically positive as it was in the context of managing the household, but it can also be reprehensible and opprobrious as it is now, in the present-day English, when applied to a woman who knows how to control other people to her advantage.
Our point is not to impose any particular solution, but merely show that building semantic word-histories on the basis of systemic/dictionary data may seem at first to be a promising methodological short-cut, but it proves in the end to be a blind alley.
The same critical note can be addressed to Górecka-Smolińska and Kleparski's (2012) examination of semantic evolution of English names of domesticated and semi-domesticated birds. Here, again, quite typically of DSP, the authors document, evidence, and discuss each of the lexemes under consideration with systemic data derived from various dictionaries and dictionary-like works. Take pheasant (ibid., pp. 226-228). 7

Pheasant
First, an etymological note on the origin of the word pheasant is offered (< Greek phasis) and supplied with a brief cultural explanation ("Georgia [w]here the pheasant is supposed to have originated and spread into the west"). What follows next is the semantic history of the word, as documented and illustrated in dictionaries, a descriptive account of the zoosemic extension pheasant has undergone since it surfaced in English in the 14 th century (from 'a well-known game-bird' via 'an article of food' to 'a victim'), and an analysis of "the historically attested pheasant-based formations" (for example, pheasant-plucker 'an unpleasant person, usually a male', or shoot a sitting pheasant 'victimize one who cannot either escape or retaliate'). Finally, the authors proceed to give "a brief account of the symbolism of the designated word". This takes us to China "where the golden pheasant was the insignia of a high-ranking civil servant", Nazi Germany where "it was a nickname for anyone who wore the golden badge of Party membership", and England where "the bird has erotic connotations". Now, in Włodarczyk-Stachurska's assessment (to appear), this account is no doubt systematical, interesting, and informative, but none of the extralinguistic facts quoted explain the unlocked zoosemic extension. That pheasant evokes erotic connotations, or that "it is the most beautiful of birds next to the peacock", or that "it was thought of a thunder bird [in China] because it was supposed to beat a drum with its wings and make thunder" simply do not relate to the from-game-bird-to-victim extension. Just to the contrary. It is difficult to see why the privileged position that pheasants seem to have been ascribed with in different cultures should go together well with meanings such as 'a victim', 'a unpleasant person', 'a split smoked herring' (kipper ), or 'a whole cold-smoked herring' (bloater ). What is questionable is not the zoosemic extension itself, but the pretense to accumulating extralinguistic facts of human cognitive and experiential assessment of the world objects (here: the pheasant) in order to motivate language change. The semantic history of pheasant is as much of a mystery before as it is after we read Górecka-Smolińska and Kleparski's account.
If an exclusively systemic approach to word-histories, as practiced in DSP, produces overgeneralizations in the case of hussy, lacks explanatory powers in the case of pheasant, it may as well miss important specific and individual developments. This we want to show now in the following section.

Maid
Kochman-Haładyj and Kleparski (2011) "propose an in-depth analysis of the semantic development and transfers of meaning content" (p. 77) of a number of historical synonyms of girl/young woman and woman in order "to trace and investigate the semantic history of [. . . ] those lexical items which have undergone the process of [. . . ] either the pejorative or the ameliorative semantic evolution" (p. 79) and those that "historically speakinghave not been subject to evaluative transformations, either pejorative or ameliorative" (p. 213). This gives us several subcategories in relation to whether or not a given word has changed its import from positive/neutral to negative (e.g., girl or lady), or from negative to positive/neutral (tabby or periwinkle), or whether it has retained its original semantic content, be it neutral (pullet), negative (dell ), or positive (job). As a result, no matter which category we take, the picture of how the examined synonyms of girl/young woman and woman have developed over the centuries is projected as exceptionally clear-cut and unequivocal.
However, this all is championed at the expense of, again, specific textual evidence, which proves especially painful if the material source being overlooked, if not just ignored, is, for example, Shakespeare's works. Any attempt at arriving at sweeping abstractions without making references to the literary giants of the Shakespeare caliber must for English be as halftrue as offering overarching generalizations for the history of Polish without referring to Mickiewicz or for the history of German without delving into Goethe.
One noteworthy example can be that of maid. Following their systemic lexicographic authorities rather selectively, Kochman-Haładyj and Kleparski (2011) present a three-fold development of the word (p. 85; notation slightly changed): < a girl, young (unmarried) woman is perceived as a maid > (13 th > 19 th centuries) < a virgin is perceived as a maid > (12 th > 19 th centuries) < a female servant is perceived as a maid > (14 th > 19 th centuries) Although all the three "perceptions" can be evidenced in/with Shakespeare's works, Kochman-Haładyj and Kleparski do not offer any quotations from Shakespeare, or, for that matter, from Elizabethan English in general.
What we lose due to this unfortunate data selection, if not textual data "circumcision", is a play upon the word maid in several of Shakespeare's plays. For example, in The Tempest, maid is used 6 times, each time in reference to Miranda, the play's only female character. Next to the three senses identified by Kochman-Haładyj and Kleparski in their dictionary search, that is 'a girl', 'a virgin', and 'a female servant', respectively, there appears yet another interesting application, of a more generic kind. 8 Having shipwrecked in a storm and made it to the shore of an unknown island, Ferdinand is separated from the other voyagers and led across the island so that he meets Miranda. Under the magic art of Miranda's father (Prospero), Ferdinand and Miranda fall in love at first sight. As expressed in Act I Scene II, she thinks he is "a spirit" or "a thing divine, for nothing natural / [she] ever saw so noble", whereas he, equally enchanted and infatuated, calls her "the goddess / on whom these airs attend" and "a wonder" only to ask this: [. . . ]  This conversation alone points to a more general meaning of maid -' a (female) human being'; Miranda is not a phantom, a spirit, or a ghost, but a human being flesh and blood. If Ferdinand did take maid here in the sense of 'a girl' or 'a virgin', he would not make this conditional marriage proposition a little while later: O, if a virgin, And your affection not gone forth, I'll make you The queen of Naples.
But the true unlocking of the meaning of maid as used in-between Miranda and Ferdinand comes somewhat later, in Act IV Scene I and Act V Scene I. In the former reference, Iris speaks of Ferdinand and Miranda as of "this man and maid", complementing the lovers for their chastity and faithfulness and presenting them as a husband and a wife, in fact. If Ferdinand is the man, Miranda is the woman.
The latter reference is even more revealing. It opens with Ferdinand's father, Alonso King of Naples, asking his son this: What is this maid with whom thou wast at play? Your eld'st acquaintance cannot be three hours: Is she the goddess that hath sever'd us, And brought us thus together?
No matter what Alonso means by maid here ('a girl' ?, 'a spirit' ?), Ferdinand finds it necessary to protest to his father's implied associations: Sir, she is mortal; But by immortal Providence she's mine.
The mortality characteristic again gives maid a generic sense of 'a real woman', or 'a true human being'.
Why should the sense of 'a (mortal) woman' be important in sketching an overall picture of maid as one of the English women terms? For two reasons. The first one is that this makes the semantic story of maid far more complex than the straightforward pejoration development postulated on the basis of the available dictionary-provided systemic data. Simply, as used in The Tempest, maid is ameliorated to the generic sense of 'a human being'. It is true that in Act III Scene I, Miranda herself diminishes her role and position to that of 'a servant', but she expresses it only to show how deeply female/human her love towards Ferdinand is: At mine unworthiness that dare not offer What I desire to give, and much less take What I shall die to want. But this is trifling; And all the more it seeks to hide itself, The bigger bulk it shows. Hence, bashful cunning! And prompt me, plain and holy innocence! I am your wife, it you will marry me; If not, I'll die your maid: to be your fellow You may deny me; but I'll be your servant, Whether you will or no. This is how Miranda sees her future in relation to Ferdinand: she can be his wife, but if rejected as his spouse ("fellow"), she is prepared to assume the role of his servant. In performing either role, she will remain truly human. Indeed, as O'Connor (2012; emphasis added) understands Miranda: The character of Miranda resolves itself into the very elements of womanhood. She is beautiful, modest, and tender, and she is these only; they comprise her whole being, external and internal. She is so perfectly unsophisticated, so delicately refined, that she is all but ethereal. [. . . ] Miranda herself appears a palpable reality, a woman, "breathing thoughtful breath," a woman, walking the earth in her mortal loveliness, with a heart as frail-strung, as passion-touched, as ever fluttered in a female bosom. [. . . ] Miranda is a consistent, natural human being. Our impression of her nymph-like beauty, her peerless grace, and purity of soul, has a distinct and individual character.
The other reason why the 'womanhood/human' characteristic cannot be forgotten is that it is only against this broader background of maid as an exceptionally tangible female human being that we can understand and explain the expressions like maid of honour 'the most important bridesmaid at a marriage ceremony', or bridesmaid 'a girl or woman who helps the bride during the marriage ceremony', or maid-in-waiting 'a woman appointed to help and attend a queen or princess'. There does not seem to be anything pejorative about any of the three. Just to the contrary. They all are widely and typically used in reference to, precisely, "a consistent, natural human being [of] a distinct and individual character" that is asked to perform her role on the basis of her strengths and virtues, such as friendship, involvement, responsibility, reliability, organizational and social skills, grace, or beauty. That is why no matter whether in the faculty of a bridesmaid, maid of honour, or maid-in-waiting, girls or women should be prepared to be anything from a facilitator via a counsellor and a friend to a partner, appearing -quite like Shakespeare's Miranda -"a palpable reality". 9

Concluding remarks
This all brings us to the question of whether we can do any better in reconstructing human conceptualizations and mentalities, as reflected in language, on the grounds of the LWV research. The answer must be a qualified 'yes'. After all, with or without "co-linguistic" evidence, the SAT strategy alone is enough to make one's research well-evidenced, well-justified, and well-balanced. In this context, Bartmiński's note on a conditional applicability of systemic kinds of data is truly revealing (2014: 284; translation is ours): 9 That being a maid in the sense of maid of honour, bridesmaid, and maid-in-waiting is more of a compliment, social distinction and elevation, rather than a chore, has recently been evoked in the context of media speculations regarding Kate Middleton's appointing her maid-in-waiting, or, to use an older and more traditional term, lady-in-waiting. This is how Malone sees the role: "The job certainly has evolved over the years. Many of the tasks that used to be allotted to ladies-in-waiting (helping her mistress dress, for example) have been assigned to other, paid members of the royal household. Today the ladies function more like social auxiliaries, helping the royal entertain dignitaries and manage her correspondence. Yet their true purpose has remained the same across the centuries: to provide appropriate companionship and wise counsel for a woman who can't exactly make friends by joining a book club and can't unwind with those friends over pints at a local pub". Or take this revealing sentence from a formal explanation on the British Monarchy's official website: "The post of Lady-in-Waiting is not open to application. In addition to The Queen, other female Members of the Royal Family appoint their own Ladies-in-Waiting" (after royal ).
As a source of "ethnolinguistically-relevant information", dictionaries are irreplaceable [Pol. nie do zastąpienia] providing (i) lexicographic definitions are critically assessed [because of their incomplete and subjective nature], (ii) differences in/among definitions are examined in a chronological order and are related to their underlying historical contexts, and (iii) the whole of the microstructure of a given entry is taken into consideration (not merely the definition itself).
This not only paves the way for textual evidence to be extensively explored and systematically used in the LWV reconstruction process, but also shows the hierarchy of available evidence, with texts being the desired source of information. Only then can browsing dictionary definitions become describing a word's semantic content, and describing a word's semantic content start turning into reconstructing the worldview behind a word's semantic content.