今日喺旺角見到呢張海報。我覺得好有創意，亦都相當完善，真係好佩服作者。但正所謂「無所因而特創者難為功，有所本而求精者易為力」，我亦都有啲感想，希望可以大家分享吓：
 |
| 「粵切字改革方案」，2020年12月16日攝於旺角 |
（1）個表入面有兩種注音嘅方法。第一種，就係用大家都好熟悉嘅拉丁字母；第二種，就係用唔同嘅字或者係邊旁去表示一啲聲母同韻母，再好似韓文咁將佢哋嵌一個方塊入面。
其實呢，如果係為咗有一個大家都接受嘅標準去表達、交流啲發音，用拉丁字母已經好足夠，亦都方便已經識得一種歐洲語言或者漢語拼音嘅人去學。
當然，如果係想順便創製一個系統去寫一啲廣東話入面而家「有音冇字」嘅字，就好似韓國嘅《訓民正音》咁，咁又唔同講法。至少，咁樣創造出來嘅方塊字，係的確比較容易融入粵文嘅。但如果係咁嘅話，或者亦都應該搵一個方法，將個聲調都放埋入個方塊裏面。
（2）其實我覺得，粵語拼音應該盡量係基於香港民間咁多年來已經形成咗嘅習慣。咁樣先至容易學，容易學先至多人學得識，先至可以普及。
（3）依家唔精確嘅嘢當然要改善：例如貝澳個「貝」同天水圍個「水」明明唔同韻，但而家好多時都係用「ui」去寫，呢啲就應該要分返開。
（4）但係，點解要用/j/去標示「央」嘅聲母？我知，粵拼（Jyutping）、國際音標（ IPA ）同埋德文都係用/j/。但如果我哋睇返個現實嘅情況，除咗乙明邨（Jat Min Chuen ）之類嘅個別例子，香港人其實係習慣用/y/嘅：例如元朗嘅元，大家都會諗到「yuen」而唔係「jyun」。
我覺得，如果已經習慣用嘅用法冇問題，就應該繼續用。所以，用/y/會好啲。
（5）如果用咗/y/做聲母，咁為咗避免混淆，或者亦都應該用個表而家冇用嘅/ue/去表示漢語拼音同德文裏面嘅「ü」，而唔係用/yu/。
咁樣仲有兩個好處：一嚟，係香港人平時都係咁用——例如「雪」，通常都會用「suet」而唔會係「syut」；二嚟，亦都有先例係咁做：例如德文打唔到ü嘅時候，就會用「ue」。
（6）因為同樣嘅道理，「z」、「c」可以用香港比較常用嘅「j」、「ch」（或者「ts」）代替。
（7）其實寫粵文真係好難，一唔小心就會夾雜咗啲白話文入去。就好似隻一直都住喺陸地上面嘅鴨，要逼自己學返識游水咁。
眾所周知，粵語歌詞要求協音(粵語所謂「啱音」)，國語卻不用。
之所以如此，我見過至少兩個解釋：
(1)比起有六調的粵語，只有四聲的國語很難配合旋律，於是便不追求協音。
(2)粵語六調裏面，只有兩個(陰上、陽上，即「九」、「五」的音)有滑音，其他的都是同一音高貫徹始終。國語的四聲，卻有三個有滑音(只有第一聲陰平不是)，所以協音無從談起。
那個才對呢？單單比較粵語和國語，是很難有結果的。
於是我想，不知道越南語的歌詞又是否追求協音呢？一方面，越南語有六調八聲，也分陰陽平上去入，可以和粵語聲調一一對應；另一方面，越南語的六調裏面，只有兩個沒有滑音：通常翻譯為「平聲」和「玄聲」、分別對應粵語的陰平、陽平的dấu ngang和dấu huyền。這樣，填詞時應該比較難和旋律協調。
 |
| 用五度標記法表達的粵語聲調。至於國語、越南語，可分別參見以上維基百科連結。 |
正所謂實踐是檢驗真理的唯一標準，我找了一首有粵、國、越三語歌詞的歌：陳奕迅的《明年今日》、《十年》，和越南歌手阮飛雄主唱的《Ngày Này Năm Sau》，從中選出二十字作比較。同樣的旋律，歌詞分別是：
粵：「明年今日 未見你一年 誰捨得改變 離開你六十年」
國：「十年之後 我們是朋友 還可以問候 只是那種溫柔」
越：「Ngày này năm sau, hẹn mà không thấy người,
Ôi đã không mong đợi,mình cũng đã khác rất nhiều」
(有文化輸入，也有文化輸出。很多廣東歌、國語歌，都有填上越南語詞，經過重新編曲，由當地歌手演繹的版本。情況有如香港八十年代為日本歌填上粵語詞：《月半小夜曲》、《千千闕歌》便是。)
為了方便比較，我把所有的音高都轉成頻率，用赫茲(Hz)表示。旋律容易處理：《明年今日》用A♭大調，上述「明」字是A♭3即207.65Hz，餘可類推。
但說話時的聲調卻是相對的：只要同一句子內每個字之間的相對音高正確，絕對音高並不重要。周星馳《食神》中，用不同語調說「乜咁啱嘅」，便是絕對音高有變，但相對音高正確的例子。填詞也是同理。側田《好人》裏面的三個「沒結果」可作參考：音高一句比一句高，但只要句子內的相對音高不變，便不會不協音。
粵語六調裡面，有四個可以是同一音高貫徹始終的。從高到低，是陰平、陰去、陽去、陽平，也就是「三」、「四」、「二」、「零」的音調；其中「三」比「四」大約高4個半音，「四」比「二」大約高3個半音，「二」比「零」又大約高2個半音。用唱名表示，大約是「la fa re do」吧。另外陰上(「九」)滑音完結時的音高大約在「三」、「四」正中間，用剛才的唱名來說則是「so」。
在五度標記法裏面，「三」、「四」、「二」、「零」、「九」的音調分別是「55」、「33」、「22」、「11」、「14」。換言之，1-5的調值可以分別轉換成「do re fa so la」。國語、越南語的音調也用五度標記法標註，也可以這樣套用：例如國語第四聲便是「51」、越南語dấu sắc(「銳聲」)是「35」等等。
把這些相對音高翻譯成以頻率表示的絕對音高，有兩種方法：
a) 把第一個字的讀音對上旋律中相對的音高。粵語中「明」是陽平、「11」、「do」，便把「1」訂為207.65Hz，其餘十九字都以此作標準。
b) 在每一句的開端，都重新和旋律作對照。例如第二句首字「未」是「陽去」，「22」、「re」，旋律中則為C4/261.63Hz，便把「未見你一年」全句都作相應調節。餘可類推。
 |
| (a) 把相對音高翻譯成以頻率表示的絕對音高，只以第一個字為準 |
 |
| (b) 把相對音高翻譯成以頻率表示的絕對音高，每句句首重設 |
把音高變化畫成圖。驟眼看來，好像比較接近國語，沒有追求協音的樣子。可是我雖然會一點國語，卻不懂越南語。為免錯誤理解維基百科中的調值，便特意問了越南來的舊同事，請把那二十字的歌詞朗讀一遍。驟耳聽來，也是覺得沒有協音。
但是他卻說有。
於是我想，一個不會越南語的人單憑肉眼讀圖，用耳朵聽，恐怕流於主觀。於是我想到一個比較客觀的方法：計算讀音與旋律頻率之間差異的標準差(standard deviation)。做法有兩種：
i) 計算每個字的讀音和對應旋律之間的頻率差異
ii) 在旋律和讀音裏面，計算每個字與前一字的頻率差異(所謂取其一階差分，first difference)，再比較兩者之間的分別。這種做法可能更能聚焦於相對音高的流轉：旋律上行時，歌詞聲調理應相應上行。
無論是哪種做法，如果旋律和讀音完全一致(例如「明年」永遠同音，「今日」永遠相差7個半音)，標準差就是0。但這是不可能的：我們討論的是歌詞，不是朗讀。
每升一個八度，頻率便增加一倍，為了方便比較，便以其對數表達頻率。
 |
| 讀音與旋律頻率之間差異的標準差 |
我的越南朋友是對的。
不過，我只是比較了二十個字，難免有以偏概全的可能。近數十年粵、國、越三語同曲異詞者甚多，相信如果能以統計方法作分析，也會相當有趣。敬以俟之來哲。
[檢疫隔離第14天，在酒店自娛時作。旨在拋磚引玉，請各方好友不吝賜教。]
===
參考資料：
I used to think Tedros Adhanom Ghebreyesus, WHO's head, had a pro-China bias. Now I think he has just been trying too hard to keep everyone happy in a futile quest to coax some of the most difficult leaders of our time into cooperating.
To that end, he had said both Trump and Xi had "done a great job". Today, it has also been revealed that WHO conspired with the Italian government to suppress a damning report on the country.
His position is not an enviable one. He
Could he have been more direct in 'handling' China, Italy, the US, etc? Probably, but it might not have been intrinsically deplorable to take a position that it's better to coax countries and keep information flowing at least to WHO officials, who could then advise other governments how lives could be saved without naming the source.
Of course, that strategy always carried a risk that he would become a useful fool to others. And, one would argue, he did.
====
There's no such thing as 'inauthentic' food: a quick survey of Hong Kong's 'pan-Western' food with examples from my quarantine meals can tell you why.
 |
| Exhibit 1: Beef spare ribs braised with red wine sauce, with a side of steamed rice, to start with a cream and butternut squash soup |
Take these beef spare ribs as an example. On the surface, the ingredients - celery, a red wine sauce - are unmistakably Western. But the side dish has been thoroughly localised.
Indeed, in most pan-Western eateries, one has exactly three side options: rice, pasta, or potatoes & salad. In that order. And by 'rice', it invariably means a portion of plain steamed jasmine rice. It's never basmati. It's never Nandos-style spicy rice either.
 |
| Exhibit 2: On the right is a localised borscht |
By convention, a local two-course pan-Western meal set starts with a soup. There are two and only two options: the 'red' soup, i.e. the house borscht, or a cream-based soup (eg Cream of Chicken / Mushroom) which is colloquially known as the 'white' soup of the day. The choice is often summarised as "red or white (soup)?" when waiters take orders. Don't confuse that with the similarly phrased question about wine.
 |
| Exhibit 3: Penne with beef brisket and shiitake mushroom |
In a similar way, Hongkongers in the UK often look down on British pan-Asian food: a kitsch mimicry of Chinese food that puts sweet and sour sauce on everything, the stereotype goes. Better stick with "authentic" Chinese restaurants and English pubs, we think (myself included).
The fashionable quest for “authenticity” could sometimes disparage hybrid cuisines as “fakes”. But they’re not. They’re fruits of cultural exchange over many years too. They may not be your taste - Wagamama has never been mine - but they’re no less “authentic”.
And we should know better.
======
If I must come up with an analogy for Hong Kong's screening system, I would say it's like a linear video game with quite a few missions that have to be cleared one by one.
 |
| First checkpoint (30.11.2020, 17:46) |
 |
| "Tel OK" |
 |
| Contents of my welcome pack |
 |
| A StayHomeSafe electronic tag |
 |
| The quarantine order station has a typically Hong Kong set-up: unpretentious, pragmatic, does the job. Almost nostalgically symbolic. (30.11.2020, 18:50) |
 |
| "BIOHAZARD": my saliva sample (30.11.2020, 19:32) |
 |
| An eerily empty airport (30.11.2020, 20:10) |
 |
| A room in the Holding Centre. There wasn't any desk so I had to have my school-meal-style dinner on top of my suitcase. (30.11.2020, 21:48) |
I suspect I've never pronounced the voiced stops /b/, /d/, and /g/ properly.
Indeed, my ears could hardly distinguish them from their voiceless counterparts /p/, /t/, and /k/ when they are unaspirated - e.g. when they're behind an /s/ in English as in "spot", "stop" and "scar" or in Romance languages like French, Italian and Portuguese.
Take the t/d set as an example. In English they correspond in fact to three sounds:
In French, Italian and Portuguese they usually correspond to #2 and #3 - a pair that’s barely distinguishable to my ears. Indeed, many of the videos I found on this topic were made to teach Spanish & Portuguese speakers how (1) in English is meant to be pronounced.
On the other hand, "T" and "D" in Cantonese and Mandarin apparently correspond to #1 and #2 above. Having grown up in a Cantonese-speaking society, perhaps I have been approximating D in English with #2 - or did I not? I cannot tell.
The reality is as long as I stick to Cantonese, Mandarin and English, I could get away with not distinguishing #2 and #3: aspiration was enough to tell "T" and "D" apart in all three of them regardless of voicedness. Indeed, this distinction would not have come to my mind if I haven't been making forays into Portuguese recently.
This probably also explains why the voiced consonants, B, D and G, are never used in Hong Kong place names - in the ears of 19th century Europeans, there were probably only different versions of Ps, Ts and Ks. For example, one could argue that Tung Chau Street has an aspirated "T" (i.e. #1), but Tung Chung has an unaspirated one (i.e. #2).
These are, of course, not the most complex Ts and Ds there could be. Hindi, for example, has eight (!) consonants that all may sound like some forms of Ts and Ds to outsiders.
(I’m sure many of you have noticed this a long time ago, but it still came as a sort of minor epiphany to me this month..)
 |
| Mapping Ts and Ds from Cantonese, English, French and Hindi. In the case of Hindi, I set out in each cell first the dental consonant, then the retroflex. In the other three, they're all somewhat alveolar. |
====
Yesterday, on the eve of a second lockdown in England, I used up my first box of masks this year.
It is one of the few boxes of EN14683 Type IIR masks I ordered on 26 January, i.e. 2nd Day of Chinese New Year. Wuhan had just gone into lockdown. Long queues formed outside Hong Kong's pharmacies, often before sunrise, for surgical masks that may or may not come on sale at 10 or 15 times their previous price. Perhaps I'll want to send some to friends & family back in Hong Kong, I thought, so I went to Boots' website and placed an order with them.
 |
| My first box of masks this year |
A few hours later I tried to place another order but it got rejected. Boots had run out of surgical masks.
When the masks arrived I posted them all save a single box of 50 to Hong Kong. Why did I do that? I certainly thought I most probably wouldn't need them, but, well, it wouldn't do any harm to keep a box on the top of my bookshelf anyway - just in case.
Surgical masks stayed hard to buy for months. We learnt casting our nets beyond the UK with the help of national registries of online pharmacies established under the European Directive 2011/62/EU (thanks), but it seems many people did as well. In the end, I only managed to get 15 more boxes from Germany, which were also airlifted to Hong Kong.
The situation was better with hand sanitisers. They had gone off the market for a week or two, but they were back by mid-February. In the few happy days when Chinese case number had started to fall, but outbreaks in Korea, Iran and Italy were still in the future, Boots even did a 'Buy 2 Get 1 Free' promotion with their pink, Valentines themed, 100mL bottles of sanitisers. I got a few dozens, kept one for myself, and posted the rest.
It had turned out to be my only bottle throughout the first lockdown.
By March the flow of masks reversed. I received two boxes of unsolicited masks from my family: one I later found out to be of the anti-dust, anti-pollution variant, and another contained 50 individually-packaged masks. "Overpackaging!", I thought, but they did come handy on my trips five months later.
In April I also received two homemade fabric from my friends in Cambridge. They have been very useful on short errand trips.
But this trusty box of Type IIR masks remained my best for a long while. Sturdy, breathable and with a strong aluminium strip that prevents steaming up when properly fitted, I saved them for my longer trips away from home. I cycled with them to Putney, to Windsor and to Cambridge. I used them on my flights to and back from Heidelberg too.
As England goes into lockdown again, it might be useful to reflect on how far we have come since the first one. Both disposable and fabric masks are now easily obtainable at reasonable prices, and I counted 70% compliance with masking in Stratford Centre the other day. Hand hygiene - and public hygiene in general - have improved. We may not be where we want to be, but we've come a long way.
And we'll be alright.
 |
| "We'll be alright", handwritten on a lunch delivery bag today. |
=========
Controlling Covid rewards politically:-
The message could not be clearer: it's not so much about left or right, global or local, climate change, immigration, tax, abortion, or China. These are serious topics - but at this moment, in many voters' minds, they rank far, far lower than staying alive and having normal lives.
And voters know if they are still living normally under the current lot's watch. As a cigar-loving, incumbent-defeating President of the United States could have said,
"It's Covid, stupid."
===
上個禮拜，好幾位臉書上嘅朋友都分享咗一篇文章，話花木蘭其實應該係蒙古人，而唔係中國人咁話喎。
其實，將現代人對民族嘅睇法套喺古人身上，好多時都會衣不稱身。東西方都係：例如喺今日法國領土上面出世、應該講古法文多過古英文嘅英格蘭國王亨利第一，係「法國人」、「諾曼人」、定係「英國人」？呢啲例子，隨手執都有幾個。
講返木蘭。女人扮男裝從軍嘅事，可能有發生過都唔奇，但《木蘭辭》就肯定係經過咗好多重嘅文學誇大同埋修飾。而故事主角嘅原型，我覺得係北魏為咗防範柔然人入侵，喺北部邊境設立嘅軍鎮嘅人。當地人嘅血緣可能係鮮卑、漢、或者混血，但自成一種講鮮卑話嘅文化。
一、現存《木蘭辭》寫低嘅時候，已經喺中原流傳多年
就我哋今日所知，最早收錄《木蘭辭》嘅係南朝陳編成嘅《古今樂錄》。但詩裏面嘅地名，即係「黃河」、「燕山」之類，都係喺塞北，似乎係北朝傳過嚟嘅詩歌。背景最有可能就係北魏同柔然人打仗嗰陣時。不過，喺南北朝，單靠地名嚟推斷未必穩陣。南朝以華夏正統自居，將好多北面嘅地名搬咗去南面，講嘢嘅時候又會用中原嘅地名借代。例如《顏氏家訓》，就話過當時南朝人「上荊州必稱陝西」；陳受梁禪嘅璽書就寫「滅陸渾於伊洛」，都只係借代啫。
不過點借代都好，首詩入面一時稱呼統治者為「天子」（「歸來見天子，天子坐明堂」，一時就稱呼佢做「可汗」（「可汗問所欲，木蘭不用尚書郎」），應該都幾肯定係北朝傳入嘅民歌，其中嘅「可汗 / 天子」就係北魏皇帝：南朝人點借代，都唔會稱呼自己嘅皇帝做「可汗」。
但依家搵到嘅記載最早係喺南朝，可能亦都意味住傳世版係喺中原傳抄，甚至口耳相傳多年嘅版本。隨住時代前進，歷代嘅演義戲曲又為傳說加添咗唔少內容。
二、文治漢化，武力鮮卑化嘅北魏政權
北魏係由拓跋部鮮卑人創立嘅。而家學界主流好似覺得鮮卑語係屬於蒙古語系，但亦都有學者覺得同突厥語系有啲淵源。點都好，都係今日蒙古高原上面嘅遊牧民族啦。北魏自從立國開始，就逐漸吸納漢人士族入朝做官，但軍隊就係以鮮卑人同埋鮮卑化咗嘅各種胡人同少數漢人為主。
北魏崛起時軍隊嘅成份，可以喺451年（北魏太平真君十二年；南朝宋元嘉二十八年）北魏太武帝圍攻盱眙城嘅時候，向宋軍守將臧質嘅信睇得出：
「吾今所遣鬥兵，盡非我國人，城東北是丁零與胡，南是氐、羌。設使丁零死，正可減常山、趙郡賊；胡死，減并州賊；氐、羌死，減關中賊。卿若殺之，無所不利。」 （《資治通鑑》第126卷）
呢度「我國人」係「我嘅國人」咁解：當時北方胡族政權都鍾意將本族人叫做「國人」。「胡」嘅狹義即係匈奴人。丁零、 氐人、羌人，都係當時中原嘅胡人。可見北魏嘅軍隊係以各種胡人為主，甚少漢人。有少少似早期羅馬嘅兵制：羅馬公民做正規軍，意大利附庸城邦嘅人就做支援兵。至於西西里、希臘行省嘅居民，一般嚟講唔參軍。
後來柔然人喺草原上面崛起，北魏就喺今日嘅內蒙古設置咗六個軍鎮（六鎮）。除咗用鮮卑本族人駐守之外，亦都徵發漢人士族子弟參軍。起初北魏首都喺平城（依家嘅大同），文武亦都冇乜分界，六鎮軍官往往可以入朝任職。但孝文帝「漢化」，將首都搬去洛陽之後，佢哋嘅地位就開始低落：
「緣邊諸鎮，控攝長遠。昔時初置，地廣人稀，或徵發中原強宗子弟，或國之肺腑，寄以爪牙。中年以來，有司乖實，號曰府戶，役同廝養，官婚班齒，致失清流。而本宗舊類，各各榮顯，顧瞻彼此，理當憤怨。」（《北史》卷五十六，《魏蘭根傳》）
佢哋對孝文帝漢化運動嘅反動，係北魏最後內戰分裂收場嘅其中一個原因。東西分裂之後，兩邊都有試圖去恢復鮮卑傳統。
三、鮮卑變成北部邊境各族嘅尚武文化認同
但去到北魏末年，所謂鮮卑，可能已經變成一個文化認同嘅象徵：一個人只要講鮮卑話，崇尚北邊尚武嘅鮮卑文化，就係鮮卑人。
後來俾後代追尊做齊高祖神武皇帝嘅高歡，佢一方面自稱關東漢人士族渤海高氏嘅後人，但同時又能夠話自己三代人喺邊疆生活，同鮮卑無分別。一啲問題都冇。呢種現象，正正顯示咗鮮卑已經唔再以血源界定。
「齊高祖神武皇帝，姓高名歡，字賀六渾，渤海蓚人也。六世祖隱，晉玄菟太守。……累世北邊，故習其俗，遂同鮮卑」（《北齊書》卷一，《神武帝紀》上）
高歡有幾鮮卑化？佢初初起兵討伐爾朱兆嘅時候，同樣係渤海高氏嘅高昂喺本鄉招咗一班漢兵。佢本能覺得漢人唔打得，提出要畀千幾個鮮卑兵比佢：
高祖曰：「高都督純將漢兒，恐不濟事，今當割鮮卑兵千餘人共相參雜，於意如何？」昂對曰：「敖曹所將部曲練習已久，前後戰鬭，不減鮮卑，今若雜之，情不相合，勝則爭功，退則推罪，願自領漢軍，不煩更配。」（《北齊書》卷二十一，《高昂傳》）
到佢喺東魏執政嘅時候，高歡就曾經好誠實咁講，鮮卑人嘅責任就係打仗，漢人嘅責任就係耕田、織布、納稅：
歡每號令軍士，常令丞相屬代郡張華原宣旨，其語鮮卑則曰：「漢民是汝奴，夫為汝耕，婦為汝織，輸汝粟帛，令汝溫飽，汝何為陵之？」其語華人則曰：「鮮卑是汝作客，得汝一斛粟、一匹絹，為汝擊賊，令汝安寧，汝何為疾之？」（《資治通鑑》第157卷）
以上種種，似乎都顯示北魏由始至終，直到520年代嘅內戰，鮮卑人同鮮卑化嘅各族軍人都係一個分開嘅武人階級。
其實呢一種兵民分治，「野蠻人」當兵，「文明人」交稅嘅社會組織形態，同差唔多時期，羅馬帝國滅亡之後喺歐洲出現嘅各日耳曼王國比較，可以話係異曲同工。
四、漢人鮮卑化，定係鮮卑人漢化？
但內戰連綿幾十年，鮮卑人死傷應該唔少。草原已經畀柔然人佔據咗，後來又落咗喺突厥人手上，唔會再有新嘅遊牧鮮卑人補充。到咗543年，喺西魏執政，後來被追尊做周太祖文皇帝嘅宇文泰，就開始要招募陝西漢人當兵：
太祖以邙山之戰，諸將失律，上表請自貶。[......] 於是廣募關隴豪右，以增軍旅。（《周書》卷二，《文帝紀》下）
但軍隊係鮮卑化、講鮮卑話嘅諗法，係唔會一時三刻就消失。西魏於是要求新嘅鮮漢混合軍隊將士──即係後來陳寅恪所謂關隴集團──都改鮮卑姓：
魏氏之初，統國三十六，大姓九十九，後多絕滅。至是，以諸將功高者為三十六國後，次功者為九十九姓後，所統軍人，亦改從其姓。（《周書》卷二，《文帝紀》下）
講開又講，高歡雖然有漢名，但係表字就用鮮卑話「賀六渾」。佢嘅對頭，西魏嘅宇文泰嘅表字就係「黑獺」。當時嘅名，好似係稱呼下級、官式場合先至會用到。同輩之間互相稱呼都係用表字。用鮮卑話起表字，有啲似我哋香港人洋名嘅用法，用法唔難理解。
到咗最後，係繼承西魏、北周嘅隋唐征服華夏。關隴集團裏面嘅家族，包括隋唐皇室，往往喺西魏、北周時期有鮮卑姓：隋文帝楊堅出世嘅時候姓普六茹，小名係梵文「那羅延」。唐高祖李淵嘅阿爺李虎，其實叫大野虎，今日嘅人可能會覺得有啲日系。直到隋朝建立，長安朝廷先至恢復漢魏衣冠，並且允許改咗鮮卑姓嘅人恢復漢姓。但去到唐朝滅亡，甚至五代十國嘅後唐，宰相名單仲係時不時會見到「豆盧」、「獨孤」之類嘅鮮卑姓。
結語
如果木蘭真係存在過，佢應該係喺北邊嘅六鎮長大，識講流利嘅鮮卑話（嗰度嘅共同語言），但亦都可能識講其他語言。以血統計，佢可能係鮮卑人、匈奴人、高車人、漢人，甚至係混血都唔出奇。
而佢會接觸到嘅同代人，對佢血統嘅興趣未必有我哋咁大。
===========
臉書連結
Discussion on Facebook in English (7 Sep 2020)
立場新聞版 (2020年10月9日)
立場新聞臉書貼文 (2020年10月9日)
 |
| 巾幗不讓鬚眉：雜貨舖入面嘅以色列女兵。Rachel Papo攝於2004年。 |
 |
| 提比里斯街景，可見歐盟旗（藍底十二星）與喬治亞國旗（五重十字）並列。自攝於2017年9月9日。 |
 |
| 提比里斯的喬治亞菜館。自攝於2017年9月9日。 |
 |
| 左起順時針：City Hall, Regent's Park, Queen Elizabeth Olympic Park |
[1] If China goes down this path and implements this national security legislation, we will be required to change the status of BN(O) passport holders and set in train arrangements which allow them to come to the UK [2] for longer than the current six-month period and to apply for extendable periods of 12-months [3] to work and study, which itself will provide a pathway to citizenship.
(the numberings are mine)
We [...] have now developed proposals for a bespoke immigration route for BNOs and their dependents. We will grant BNOs five years’ limited leave to remain, with a right to work or study. After these five years, they will be able to apply for settled status, and after a further 12 months with settled status, they will be able to apply for citizenship. This is a special, bespoke set of arrangements developed for the unique circumstances we face and in the light of our historic commitment to the people of Hong Kong.
All those with BNO status will be eligible, as will their family dependents who are usually resident in Hong Kong, and the Home Office will put in place a simple, streamlined application process. I can reassure hon. Members that there will be no quotas on numbers.This offer is slightly more generous than what was announced on 28 May 2020, but details remain scanty. Most observations above except (1) and (2) remain relevant.
 |
| 在意大利巴里(Bari)偶遇的天主教巡遊，自攝於2012年6月10日 |
 |
| 根據Youtuber Ollie Bye製作的影片，永寧寺矗立時為世界第二高建築物，僅次於埃及大金字塔。 |
「肅初入國，不食羊肉及酪漿等物，常飯鯽魚羹，渴飲茗汁。京師士子道肅一飲一斗，號為漏卮。經數年已後，肅與高祖殿會，食羊肉酪粥甚多。」對於褒北貶南，楊氏有時好像太過用心。書中記城東景寧寺事時，便引用帶兵護送元顥入洛的南朝梁將領陳慶之說：
「自晉宋以來，號洛陽為荒土，此中謂長江以北盡是夷狄。昨至洛陽，始知衣冠士族並在中原。禮儀富盛，人物殷阜，目所不識，口不能傳。所謂帝京翼翼，四方之則，如登泰山者卑培塿，涉江海者小湘沅。北人安可不重？」這段話好像不時會用來引證「孝文帝遷都後，北朝勝於南朝」。按陳慶之在洛陽時為永安二年（529年）五月。而楊氏記城內永寧寺事時，就說去年（528 / 建義元年）四月十三日河陰之變「新經大兵，人物殲盡」。不知陳慶之所見留在城中，打算依附北海王的「衣冠士族」、殷阜人物有多少？且別說陳慶之此後便回到南方，在河北的作者所聽聞的語錄，又有多可信了。
 |
| 自製的PDF書 |
「1900至45年日本為中國帶來的利多於弊」你是否同意此說？試參考資料C及D，並就你所知，解釋你的答案。(8分)
其中資料C是1905年日本法政大學校長梅謙次郎致清廷的文章，支持中日教育合作；資料D是1912年1月黃興寫給日本政客井上馨的信，以及中華民國臨時政府與三井財閥於1912年2月簽訂的一份合同。內容是日本借款支持民國革命政府之類。詳情可參考明報報導
甲：「閼逢」(爾雅)；「焉逢」(史記)《淮南子》都有，但我見基本上都係不出《爾雅》、《史記》範圍。
乙：「旃蒙」(爾雅)；「端蒙」(史記)
寅：「攝提格」(爾雅)；「攝提格」(史記) [一樣]
酉：「作噩」(爾雅)；「作鄂」(史記)
蘆葉滿汀洲。寒沙帶淺流。二十年、重過南樓。我說，載酒這個古今都不可或缺，但買桂花幹甚？又不能佐酒。往昔與朋友泛舟劍川，往往是在瑪莎、升記買紅白酒，佐以煙肉、煙魚、薯片等物，但從來沒有買過花。當然，如果「桂花」是杞子桂花糕的代稱，那麼配上葡萄美酒夜光杯，或者黃酒大閘蟹，也是好的。
柳下系舟猶未穩，能幾日、又中秋。
黃鶴斷磯頭。故人今在否。舊江山、渾是新愁。
欲買桂花同載酒，終不似、少年遊。
鍾陵醉別十餘春，重見雲英掌上身。「可能」和「俱是」。
我未成名卿未嫁，可能俱是不如人。
 |
| 圖文不符：鄧寇克海灘，攝於2019年7月27日，17:36 |
 |
| 威川。左方隱約可見的城堡、教堂，即對衡城中心。攝於2013年6月7日 16:13 |
 |
| 在路上與行裝合影，2013年6月9日 19:27 |
 |
| 下山時所見雲隙光，攝於2013年6月10日 18:41 |
 |
| 給自己看的旅行日誌 |
 |
| Bloomberg / Getty as used by The Atlantic |
 |
| A mask-wearer at a London airport. Source: Sky |
 |
| Well-stocked Sainsbury's in Stratford Centre, morning of 4 March 2020. Hopefully this will last... |
