解开“主动词汇”、“被动词汇”之谜的新思路（英语教学法原著选读72）

以下英文原文节选自上海外语教育出版社出版的《教学法丛书：词汇：描述、习得与教学》一书。

导读

前两周我们读了Francine Melka的论文“Receptive and Productive Aspects of Vocabulary”，了解到所谓“主动词汇”、“被动词汇”，很可能只是人们直觉上的一种错误，因为测量工具的局限与误差、各种语言因素、非语言因素对这两种“词汇量”的测量影响很大。Melka更进一步提出了新的看法。本周我们来看看另一位学者Paul Meara提出的一种新的研究模型，这种思路对解开“主动词汇”、“被动词汇”之谜不无帮助。这种模型认为，词汇的“习得事件”（acquisition event）包括了一种“新碰到的词和已经掌握的词之间建立的联系”。这种联系可以是双向的（比如看到A想起B，看到B也能想起A），也可能是单向的（比如看到A能够想起B，但看到B不一定想起A）。那么一般来说，新词和已经学过的词之间建立的联系都是单向的，所以这时已经学过的词之间发生的语义联系就不会牵涉到新词，比如建立了A到B的联系，但没有建立B到A的联系，则C、D、E和B的词义互动是不会对A的掌握有什么作用的。这时，A（新学的词）就很像是一般概念中的“被动词汇”，如果此后与A的接触仅限于A到B的联系，则A无法变成主动词汇，无法和其他词建立联系。

那么，被动词汇怎样变成主动词汇呢？根据Meara所提出的研究模型，并不是由于见得多了，而是由于出现了新的情景——新的外部刺激，建立了B到A乃至C到A的联系，这样一来A就进入了B、C、D、E等已掌握的词构成的圈子，一跃而成为“主动词汇”，因为这时它就能和其他词汇发生互动了。

虽然下列选文接下来说“这个模型明显是有缺陷的”，但我觉得这个模型基本上解释了主动词汇和被动词汇之间的种种关系，已经是很有用的了。说起来也有意思，Meara这篇论文“Towards a new approach to modeling vocabulary acquisition”本意是建议应用语言学家要多采用数学模型——或者，至少是语言模型——来研究语言，就像心理学家们研究语言那样，我却从中找到了一个解决主动词汇、被动词汇之谜的新思路。当年美国作家Upton Sinclair的作品《屠场（The Jungle）》揭露了美国屠宰行业的黑暗，控诉的是美国资本主义社会的罪恶，却唤醒了美国公众对食品安全立法的重视，和他的本意几乎是完全不相干了。有评论家评论说，Sinclair是“瞄准了心，却打中了胃”，我觉得这次Meara也有点像。

最后，很多朋友肯定希望我举个例子说明一下Meara的这个模型，那我就来说说。我曾有一次给一个新教师职前培训班上课，我用的是英语讲课，其中有一句话说，I profess that I agree to the views stated in the book completely.就有旁听的老师表示惊讶，他们说以前只知道professor，当然从processor到process的变化也知道有profess这么个词，但从来没想过profess是这么用的。我相信我这个句子对他们来说就是一个外部刺激，帮助他们建立了profess和that从句之间的关系，从那以后profess对他们来说很可能就不再是一个被动词汇，而变成主动词汇了。

原文 

Let us suggest, for the purposes of an initial model, that an acquisition event consists of the building of a connection between a newly encountered word, and a word that already exists in the learner’s lexicon. This connection might be a link between the new L2 word and its L1 translation equivalent, or it might be a link between the new L2 word and an already known L2 word. We also need to specify whether the links are bidirectional (in the sense that a link from A to B is also a link from B to A), or just unidirectional (in the sense that a link from A to B does not necessarily imply a link from B to A). ...

The main advantage of our model is that it allows us to think about vocabulary acquisition as a cumulative activity, rather than an all-or-nothing affair. Unknown words are words that have no connections of any kind to the learner’s lexicon. Known words are words that are connected, but the number of these connections may vary. This gives us a natural mechanism for talking about words which are ‘known’ to a greater or lesser extent. Poorly-known words are words with few connections, while better-known words are simply words with many connections. On this model, any word which which is encountered frequently enough will, in time, develop a rich set of connections with other words. ...

More importantly, perhaps, the model suggests that some random acquisition events may be particularly important. Consider the situation shown in Figure 2. This figure shows a word (W) which has some limited connections to the main body of the lexicon (L), but all these connections flow in one direction, from W(the word) to L(the main lexicon). There is no pattern of connections running in the opposite direction. Clearly, a word like W is anomalous, in that it is part of the lexicon in only a limited sense. It can affect L, in the sense that activation can spread from W to L, but W is not itself affected by anything that happens in L. How important this is depends on exactly what we think the nature of the links joining words in the lexicon might be. For example, if we think of the links as connections that allow activity in one part of the lexicon to spread to other parts of the lexicon, then the pattern of connections enjoyed by W means that it cannot share in the general activation patterns in L. On the other hand, if W could be activated by an external stimulus, then its activation could be passed on to other parts of the lexicon. This description of W as an item that can be activated by an external stimulus, but not activated spontaneously, makes it sound suspiciously like a passive/receptive vocabulary item. This in turn suggests that the crucial distinction between active and passive vocabulary might simply be that active vocabulary items are connected to their parent lexicons by more than one type of connection. Clearly, if this idea is correct, then the encounter situation that allows the first connection from L to W to be established will have the effect of turning W from a passive/receptive item to an active/productive one. This is a rather different way of thinking about the way passive vocabulary becomes activated than what we have encountered so far. In particular, it implies that a new item can become active as a result of a single exposure, but it also implies that there is no natural progression from a passive state to an active one. It also implies that being active might be a transitory state, depending on which other parts of the lexicon are activated. If W is connected to a part of L which is not activated, then there is no activation to spread, and W will remain in a passive state.